Full text of "Radio proximity fuzes for fin-stabilized missiles"

See other formats

LC REGUT, A
OR REP
DOCtP

MARK

t>t~ "IE SERVICING
'ART OF THIS
ICATION
2 AJCELLED.

«3

ts 8?

€B a

c;2 §

• br. «-

S <• o

> UM >s
- *- ■ 1 S3

r,i 5

Sag

Ssg

SKI

RECLASSIFIED
By authority Secretary of

SFP 1 1960
Defense memo 2 August I960
LIBRARY OF CONGRESS

a ^sr

a$>

*$y a ;%."•• /f*

& ch^^s*

^ /S v>- ryr

; 4'

^ A

A/v

;Fy>

^ ; V ^

SUMMARY TECHNICAL REPORT

OF THE

NATIONAL DEFENSE RESEARCH COMMITTEE

This document contains information affecting the national defense of
the United States within the meaning of the Espionage Act, 50 U. S. C.,
31 and 32, as amended. Its transmission or the revelation of its con-
tents in any manner to an unauthorized person is prohibited by law.

This volume is classified^^^|^^B in accordance with security regu-
lations of the War and N av^^^partments because certain chapters
contain material which was (SEGSH at the date of printing. Other
chapters may have had a low^WBIBfication or none. The reader is
advised to consult the War and Navy agencies listed on the reverse
of this page for the current classification of any material.

Manuscript and illustrations for this volume were prepared
for publication by the Summary Reports Group of the
Columbia University Division of War Research under con-
tract OEMsr-1131 with the Office of Scientific Research and
Development. This volume was printed and bound by the
Columbia University Press.

Distribution of the Summary Technical Report of NDRC
has been made by the War and Navy Departments. Inquiries
concerning the availability and distribution of the Summary
Technical Report volumes and microfilmed and other refer-
ence material should be addressed to the War Department
Library, Room 1A-522, The Pentagon, Washington 25, D. C.,
or to the Office of Naval Research, Navy Department, Atten-
tion : Reports and Documents Section, Washington 25, D. C.

Copy No.

119

This volume, like the seventy others of the Summary Tech-
nical Report of NDRC, has been written, edited, and printed
under great pressure. Inevitably there are errors which have
slipped past Division readers and proofreaders. There may
be errors of fact not known at time of printing. The author
has not been able to follow through his writing to the final
page proof.

Please report errors to :

JOINT RESEARCH AND DEVELOPMENT BOARD
PROGRAMS DIVISION (STR ERRATA)

WASHINGTON 25, D. C.

A master errata sheet will be compiled from these reports
and sent to recipients of the volume. Your help will make
this book more useful to other readers and will be of great
value in preparing any revisions.

SUMMARY TECHNICAL REPORT OF DIVISION 4, NDRC

VOLUME 1

RADIO PROXIMITY FUZES
FOR FIN-STABILIZED
MISSILES

OFFICE OF SCIENTIFIC RESEARCH AND DEVELOPMENT
VANNEVAR BUSH, DIRECTOR

NATIONAL DEFENSE RESEARCH COMMITTEE
JAMES B. CONANT, CHAIRMAN

DIVISION 4

ALEXANDER ELLETT, CHIEF

WASHINGTON, D. C., 1946

RET

NATIONAL DEFENSE RESEARCH COMMITTEE

James B. Conant, Chairman
Richard C. Tolman, Vice Chairman
Roger Adams Army Representative1

Frank B. Jewett Navy Representative2

Karl T. Compton Commissioner of Patents3

Irvin Stewart, Executive Secretary

1 Army representatives in order of service :

Maj. Gen. G. V. Strong
Maj. Gen. R. C. Moore
Maj. Gen. C. C. Williams
Brig. Gen. W. A. Wood, Jr.

Col. E. A.

Col. L. A. Denson
Col. P. R. Faymonville
Brig. Gen. E. A. Regnier
Col. M. M. Irvine
Routheau

2 Navy representatives in order of service :

Rear Adm. H. G. Bowen Rear Adm. J. A. Furer
Capt. Lybrand P. Smith Rear Adm. A. H. Van Keuren
Commodore H. A. Schade

3 Commissioners of Patents in order of service :
Conway P. Coe Casper W. Ooms

NOTES ON THE ORGANIZATION OF NDRC

The duties of the National Defense Research Committee
were (1) to recommend to the Director of OSRD suit-
able projects and research programs on the instru-
mentalities of warfare, together with contract facilities
for carrying out these projects and programs, and (2)
to administer the technical and scientific work of the
contracts. More specifically, NDRC functioned by initi-
ating research projects on requests from the Army or
the Navy, or on requests from an allied government
transmitted through the Liaison Office of OSRD, or on
its own considered initiative as a result of the expe-
rience of its members. Proposals prepared by the Divi-
sion, Panel, or Committee for research contracts for
performance of the work involved in such projects were
first reviewed by NDRC, and if approved, recommended
to the Director of OSRD. Upon approval of a proposal
by the Director, a contract permitting maximum flexi-
bility of scientific effort was arranged. The business
aspects of the contract, including such matters as mate-
rials, clearances, vouchers, patents, priorities, legal
matters, and administration of patent matters were
handled by the Executive Secretary of OSRD.

Originally NDRC administered its work through five
divisions, each headed by one of the NDRC members.

These were:

Division A — Armor and Ordnance
Division B — Bombs, Fuels, Gases, & Chemical Problems
Division C — Communication and Transportation
Division D — Detection, Controls, and Instruments
Division E — Patents and Inventions

In a reorganization in the fall of 1942, twenty-three
administrative divisions, panels, or committees were
created, each with a chief selected on the basis of his
outstanding work in the particular field. The NDRC
members then became a* reviewing and advisory group
to the Director of OSRD. The final organization was as
follows :

Division 1 — Ballistic Research

Division 2 — Effects of Impact and Explosion

Division 3 — Rocket Ordnance

Division 4 — Ordnance Accessories

Division 5 — New Missiles

Division 6 — Sub-Surface Warfare

Division 7 — Fire Control

Division 8 — Explosives

Division 9 — Chemistry

Division 10 — Absorbents and Aerosols

Division 11 — Chemical Engineering

Division 12 — Transportation

Division 13 — Electrical Communication

Division 14 — Radar

Division 15 — Radio Coordination

Division 16 — Optics and Camouflage

Division 17 — Physics

Division 18 — War Metallurgy

Division 19 — Miscellaneous

Applied Mathematics Panel

Applied Psychology Panel

Committee on Propagation

Tropical Deterioration Administrative Committee

Library of Congress

201 5 490929

NDRC FOREWORD

AS events of the years preceding 1940 re-
vealed more and more clearly the serious-
ness of the world situation, many scientists in
this country came to realize the need of organ-
izing scientific research for service in a national
emergency. Recommendations which they made
to the White House were given careful and
sympathetic attention, and as a result the Na-
tional Defense Research Committee [NDRC]
was formed by Executive Order of the Presi-
dent in the summer of 1940. The members of
NDRC, appointed by the President, were in-
structed to supplement the work of the Army
and the Navy in the development of the instru-
mentalities of war. A year later, upon the estab-
lishment of the Office of Scientific Research and
Development [OSRD], NDRC became one of
its units.

The Summary Technical Report of NDRC is
a conscientious effort on the part of NDRC to
summarize and evaluate its work and to present
it in a useful and permanent form. It com-
prises some seventy volumes broken into groups
corresponding to the NDRC Divisions, Panels,
and Committees.

The Summary Technical Report of each Di-
vision, Panel, or Committee is an integral sur-
vey of the work of that group. The report of
each group contains a summary of the report,
stating the problems presented and the philos-
ophy of attacking them, and summarizing the
results of the research, development, and train-
ing activities undertaken. Some volumes may be
“state of the art” treatises covering subjects to
which various research groups have contrib-
uted information. Others may contain descrip-
tions of devices developed in the laboratories. A
master index of all these divisional, panel, and
committee reports which together constitute the
Summary Technical Report of NDRC is con-
tained in a separate volume, which also includes
the index of a microfilm record of pertinent
technical laboratory reports and reference ma-
terial.

Some of the NDRC-sponsored researches
which had been declassified by the end of 1945
were of sufficient popular interest that it was
found desirable to report them in the form of
monographs, such as the series on radar by
Division 14 and the monograph on sampling in-
spection by the Applied Mathematics Panel.
Since the material treated in them is not dupli-
cated in the Summary Technical Report of
NDRC, the monographs are an important part

of the story of these aspects of NDRC research.

In contrast to the information on radar,
which is of widespread interest and much of
which is released to the public, the research on
subsurface warfare is largely classified and is
of general interest to a more restricted group.
As a consequence, the report of Division 6 is
found almost entirely in its Summary Technical
Report, which runs to over twenty volumes. The
extent of the work of a Division cannot there-
fore be judged solely by the number of volumes
devoted to it in the Summary Technical Report
of NDRC ; account must be taken of the mono-
graphs and available reports published else-
where.

The program of Division 4 in the field of elec-
tronic ordnance provides an excellent example
of the manner in which research and develop-
ment work by a civilian technical group can
complement and supplement work done by the
Armed Services. The greatest responsibility of
Division 4, under the leadership of Alexander
Ellett, was to undertake the development of
proximity fuzes for nonrotating or fin-stabilized
missiles, such as bombs, rockets, and mortar
shells.

Early work on fuzes of various types indi-
cated that those operating through the use of
electromagnetic waves offered the most promise ;
the eventual device depended on the doppler
effect, combining the transmitted and received
signals to create a low frequency beat which
triggered an electronic switch. During the last
phases of the war against Japan, approximately
one-third of all the bomb fuzes used by carrier-
based aircraft were proximity fuzes. For im-
proving the accuracy of bombing operations,
the Division developed the toss bombing tech-
nique, by which the effect of gravity on the
flight path of the missile is estimated and
allowed for. The success of this technique is
demonstrated by its combat use, when a circle
of probable error as low as 150 feet was
obtained.

The Summary Technical Report of Division
4 was prepared under the direction of the Di-
vision Chief and has been authorized by him for
publication. We wish to pay tribute to the enter-
prise and energy of the members of the Di-
vision, who worked so devotedly for its success.

Vannevar Bush, Director
Office of Scientific Research and Development
J. B. Conant, Chairman
National Defense Research Committee

FOREWORD

The primary program of Division 4, NDRC,
was development of proximity fuzes for
bombs, rockets, and trench mortar projectiles.
The National Bureau of Standards [NBS] pro-
vided facilities and personnel for the Division’s
Central Laboratory and the Division (or its
predecessor, Section E of Division A) served
as the principal liaison between NDRC and
NBS. In large measure the developments pre-
sented in this Division 4 STR must be credited
to the National Bureau of Standards. Credit
also is due the Ordnance Department of the
Army for excellent cooperation. The main-
tenance of effective liaison was due largely to
Colonel H. S. Morton, whose enthusiasm for
the program coupled with intelligent criticism
and suggestions based on sound technical
knowledge contributed much of value.

The present volume summarizes the Divi-
sion’s development of radio proximity fuzes.
The technical direction of this development was
throughout in the able hands of Harry Dia-
mond, leader of the little radio fuze group
organized at the Bureau of Standards in De-
cember 1940, and finally Chief of the Bureau’s
Ordnance Development Division. Throughout
the program, he received invaluable technical
assistance from W. S. Hinman, Jr., Chief
Engineer of the aforementioned NBS division.
The excellent presentation found here is due
to the editor of these three volumes, A. V.
Astin, Assistant Chief of the Ordnance De-
velopment Division, NBS.

Other Division 4 contractors made valuable
contributions to particular projects on which
they were engaged. Deserving of special men-
tion are the University of Florida for work on
trench mortar fuzes, the Globe-Union Com-
pany of Milwaukee for work on safety and
arming devices and ceramic circuits, and the
University of Iowa for improved recovery de-
vices and a smooth working proof organization.
The development of generator power supplies
was largely carried out by the Westinghouse
Company in Baltimore and by the Zenith Radio
Corporation.

Reliability of radio fuzes depends at least as
much on good production methods and tech-
niques as upon good design. In the solution of
production problems outstanding contributions
were made by the Zell Corporation, Baltimore,
and Bowen and Company, Bethesda, Maryland,
who operated pilot lines; and by the Arnold
Engineering Company, the Emerson Radio and
Phonograph Corporation, the General Electric
Company, the Globe-Union Company, the
Philco Corporation, the Raytheon Manufactur-
ing Company, the Sylvania Electric Products,
Inc., the Westinghouse Electric and Manufac-
turing Company, the Rudolph Wurlitzer Com-
pany, and the Zenith Radio Corporation, who
produced fuzes or fuze components.

Alexander Ellett
Chief, Division 4

SECRET

vii

PREFACE

The Summary Technical Report of Division
4 has been prepared in three volumes :
Volume 1, describing the work on radio prox-
imity fuzes, the major work of the division;
Volume 2, discussing bomb, rocket, and torpedo
tossing, a new fire control method for airborne
missiles; and Volume 3, a report on various
miscellaneous projects. An overall summary of
the Division 4 program appears as Chapter 1 in
Volume 3.

The present volume treats the technical prob-
lems relating to the design, production, and use
of radio proximity fuzes for fin-stabilized (non-
rotating) missiles, including bombs, rockets,
and trench mortar shells. For a treatment of
work on fuzes for spin-stabilized projectiles, the
reader is referred to the reports of Section T
of OSRD. For work on other types of proximity
fuzes for fin-stabilized missiles, the reader is
referred to Volume 3 of the Division 4 STR.
The latter reference includes a general survey
of various types of proximity fuzes and a de-
tailed summary of the work done by Division 4
on photoelectric fuzes.

A primary consideration in the preparation
of this volume has been to arrange the material
so that it will be useful for reference purposes.
To fulfill this objective, the various chapters are
reasonably self-contained, and each chapter
may be read separately without too much loss
in meaning. This mode of presentation has, of
course, resulted in some duplication of ma-
terial, but it is believed that the advantages
justify the extra space required. Numerous
cross references between the chapters are in-
cluded to facilitate expansion or clarification of
various items.

For the reader who is interested primarily
in the essential operating characteristics of the
radio proximity fuzes placed in production,
Chapter 5, “Catalogue of Fuze Types,” is the
only part of this volume which need be read.
The catalogue chapter also includes a descrip-
tion of the important features of design for
each of the various fuzes.

The introduction to the volume (Chapter 1)
explains the objectives of the development pro-
gram, how radio fuzes operate, and includes a

brief summary of the accomplishments in the
development and production program.

Chapter 2 discusses in detail the basic theory
of operation and shows how the required
operating characteristics of a fuze may be con-
verted into an engineering design problem. The
material of Chapter 2 is fundamental to any
fuze design involving interaction of radio waves
with the target. Because of the great potential
use of this theory in future development work,
the treatment of Chapter 2 is much more
thorough than would appear necessary merely
as a summary of completed work.

The methods by which the electrical design
problems were solved are discussed in Chapter
3. Section 3.4 of Chapter 3 deals with the de-
sign of generator power supplies, one of the
outstanding features of the later fuzes de-
veloped by Division 4. Although this section is
included in the electrical design chapter, it
contains considerable material relating to the
mechanical design of generators. A clear-cut
separation of the mechanical and electrical de-
sign requirements for the generator was not
practicable. Chapters 2 and 3 are quite technical
in nature and will probably be of interest only
to scientists and engineers. These chapters may
be omitted by the nontechnical reader.

Chapter 4 analyzes the problems of mechan-
ical design and layout and includes a treatment
of the arming and safety features of the
fuzes.

Chapter 6 describes the production methods
and summarizes accomplishment in the produc-
tion program. Since the problems of reducing a
laboratory design of a proximity fuze to a model
which could be built in mass production were
fundamental to the entire program, the story
of this chapter is of basic importance. It should
be of interest to both the technical and the non-
technical reader.

Chapters 7 and 8 describe the methods of
testing proximity fuzes in order that their
quality might be evaluated and their perform-
ance under operational conditions predicted.
The former chapter is concerned with labora-
tory test methods and quality control. A de-
scription of testing apparatus is included. The

PREFACE

latter chapter deals with field test methods and
proving ground procedures in which opera-
tional conditions were simulated.

Chapter 9 gives a somewhat more detailed
analysis of the operating characteristics of the
fuzes than is given in Chapter 5 in that the
results of all important tests which were car-
ried out on the fuzes are summarized. The
chapter includes evaluations of performance
for each of the fuze types under a variety of
operating conditions. The operational experi-
ence is also presented in this chapter.

An analysis of problems pertaining to
countermeasures and counter-countermeasures
has not been included in this volume.

The successful development of radio prox-
imity fuzes, or VT fuzes as they are commonly
called, involved the cooperative efforts of many
organizations and individuals. A listing of all
of the individuals who contributed to the suc-
cess of the program would be an extremely
difficult, perhaps even impossible, task. How-
ever, the organizations which participated in
the development program are listed at the end
of the volume.

This volume was prepared by the staff of the
Ordnance Development Division of the Na-
tional Bureau of Standards, which served as
the central laboratories for Division 4. Reports
of the various contractors to Division 4 have

been used freely, and these are listed as refer-
ences in the bibliography.

The editor wishes to take this opportunity to
record thanks and appreciation for the efforts
of the many individuals who cooperated in the
preparation of the volume. In particular, some
of these are: Dr. Robert D. Huntoon, who as-
sisted in the overall planning of the volume and
who was also the senior author of Chapter 2;
Dr. Alexander Ellett and Mr. Harry Diamond,
Chief, Division 4 and Chief, Ordnance Develop-
ment Division, respectively, who offered valu-
able suggestions and advice on numerous items ;
other authors who are listed in the table of
contents as well as in footnotes to the various
sections which they prepared ; Mr. Theodore C.
Hellmers, who prepared the photographs used
in this volume (unless credit is otherwise in-
dicated) ; Mr. E. W. Hunt and his staff for their
diligent and painstaking efforts in the prepara-
tion of other art work; Miss Lee Smolen and
Mrs. Henrietta Leiner for preparation of
bibliographical material; and Miss Helen Olm-
stead, Mrs. Betty Hallman, and Miss Jane
Grant for their untiring efforts in the prepara-
tion, assembly, and correction of manuscripts.

A. V. Astin
Editor

CONTENTS

CHAPTER PAGE

1 Introduction 1

2 The Radiation Interaction System 17

3 Electronic Control Systems 81

4 Mechanical Design 167

5 Catalogue of Fuze Types 209

6 Production 245

7 Laboratory Testing of Fuzes 278

8 Field Testing of Proximity Fuzes 312

9 Analysis of Performance 360

Glossary 433

Bibliography 437

OSRD Appointees 463

Contract Numbers 464

Service Project Numbers 467

Index 469

Strike photograph of the first operational use of proximity fuzed bombs. The target is the beach area of
I wo Jima during the pre-invasion bombing of the island. The characteristic crescent-shaped fragmentation
patterns of air-burst bombs are clearly recognizable. (Army Air Force photograph.)

t,C RE'--'
OR RT’"
DOCTT
MARK J

rT BEFORE SERVICING
T NG AM PART OF THIS
V , C Ac T TCATION
.iU3T BE _CA T3E ,LEDT

Chapter 1

INTRODUCTION

LC REGULATION: BEFORE SERVICING
OR REPRODUCING ANY PART OF THIS
DOCUMENT, ALLjCLASSIFICATION
MARKINGS MUST BE CANCELLED:

1 * OBJECTIVES AND MILITARY
REQUIREMENTS

Radio proximity fuzes are intended to deto-
nate missiles automatically upon approach
to a target and at such a position along the
flight path of the missile as to inflict maximum
damage to the target.

The optimum position for detonation of the
missile depends upon the nature of the target
and the properties of the missile. Conditions
of use divide possible targets into two major
groups: (1) airborne targets, and (2) surface
targets either on the ground or on water. These
two applications are referred to variously as
(1) antiaircraft, air-to-air, ground-to-air, and
(2) ground-approach, air burst, air-to-ground,
ground-to-ground.

As a class, proximity fuzes belong with time
fuzes, in contrast with contact fuzes, since they
are useful wherever contact of the missile with
the target or penetration into the target is not
necessary to inflict damage. Because of the au-
tomatically accurate nature of their operation,
proximity fuzes not only extensively replace
time fuzes, but they make possible many new
and important applications for which time fuzes
would be ineffective. They also replace contact
fuzes in many applications where contact with
an object, not necessarily the target, is used
merely as a triggering operation for the fuzes
and not because contact is essential to inflict
damage.

Military requirements for proximity fuzes
became specific and well defined only after the
development had passed the exploratory stage.
Initially the requirements were quite general;
(1) the fuze should detonate the missile “in the
vicinity” of the target, (2) the fuze should be
as small and rugged as possible, (3) it should
be safe for handling and operational use, (4) it
should perform reliably under a wide range of
service conditions, (5) it should require a mini-
mum of special equipment and training for its
operational use, (6) it should be relatively im-
mune to possible enemy countermeasures, and

(7) in antiaircraft weapons, it should have a
self-destruction [SD] feature to operate, in case
of a miss, after passing the target. Most of the
foregoing requirements could not be more ac-
curately specified until a certain amount of
design experience was available or until actual
fuzes were available for proving ground tests.

For example, the careful definition of the
proper point on the trajectory for the fuze to
function had to be based on experimental trials
using fuzes against actual or simulated targets.
Before the fuzes could be built for such tests,
estimates were required concerning the ex-
pected optimum conditions. In the antiaircraft
case, it was fairly obvious that the position of
function should be matched to the dynamic frag-
mentation pattern of the missile so that the
greatest number of fragments would be di-
rected at the target. To achieve the proper
directional sensitivity, a number of factors, as
shown in Chapters 2 and 3, had to be balanced
against each other, and the final specification
of performance was based on numerous design
compromises and field tests. In the ground tar-
get case, no experimentally verified optimum
burst heights were available until the end of
1944 and then only for limited types of missiles
and targets/ For many important ground tar-
get applications, optimum burst heights are still
undetermined.

Some of the mechanical features were capa-
ble of more exact specification. Although small
size and ruggedness were objectives toward
which improvement was continuous, certain
minimum requirements were definite very early
in the program. Bomb fuzes were to be bal-
listically interchangeable with regular fuzes so
that their use would require no modifications
in bombing tables. Available stowage space in
bomb bays made it necessary to impose limita-
tions on overall length, and a maximum exten-
sion of 5 in. beyond the nose of the bomb was
prescribed, although shorter fuzes were pre-

a These statements refer specifical^~^G tfie ^uzes fur"
fin -stabilized or nonrotating nB§Bila*lth<Mlit$s SftGM&fc^ry of
rockets, and trench-mortar shells.

SEP 1 196p

Defense memo 2 August 1960

LIBRARY OF CONGRESS

INTRODUCTION

ferred. Standard fuze-well cavities in bombs
fixed other dimensions. A minimum require-
ment on ruggedness was that the fuze with-
stand any vibrations or accelerations of the
missile. There were also standard military
rough-handling specifications but these were
more of a requirement for packaging than for
fuze design.

The arming and safety requirements, with
one important exception, had to be worked out
experimentally as the development progressed.
The exception was the specification for an inter-
rupted powder train between the detonator and
booster, a standard Army Ordnance technique
which was required of all proximity fuzes.
Since proximity fuzes are, by their very nature,
susceptible to their surroundings and unable
to distinguish between friendly and enemy tar-
gets, the arming problem is appreciably differ-
ent than with ordinary fuzes. In general, longer
“safe” times after firing or release of the mis-
sile are desired for proximity fuzes, but an
ideal safe period compromises the usefulness of
the weapon. The details of the development of
the arming and safety features and require-
ments are discussed in Chapter 4.

The very necessary exploratory work on
radio proximity fuzes, was done under rather
general requests from the Services, including
a conference on August 12, 1940, between rep-
resentatives of the Navy Bureau of Ordnance
and NDRC j1 Projects OD-27, dated January 14,
1941, and OD-3B, dated June 11, 1941, of the
Army Ordnance Department; and Project
CWS-19, dated August 30, 1941, from the
Chemical Warfare Service. The pertinent mili-
tary characteristics for fuzes covered by these
authorizations were essentially as outlined.

After laboratory development and field tests
had established general possibilities and limits
for radio proximity fuzes, specific Service re-
quirements were put forth based on anticipated
operational needs. The first major project which
was carried through to large-scale production
was for the T-5 fuze to be used with the Army’s
4^4-in. (M-8) rocket. The desired characteris-
tics for this fuze2’ 3 were, in addition to the
general requirements stated above, (1) the
complete fuze should fit into a cylindrical con-
tainer approximately 2% in. in diameter and

5 in. long, with an allowable conical extension
on the front end of the cylinder about 2 in.;

(2) at least 50 per cent of the fuzes were re-
quired to function in the vicinity of an airplane
target when fired on the rocket and within the
lethal range of the fragments of the rocket;

(3) the fuze was to be armed and operative
approximately V2 sec after firing; and (4) the
fuze should have an SD element operating ap-
proximately 9 sec after firing.

The T-5 fuze project was limited in that the
intended use was confined to a single missile
and for a single application, antiaircraft. It
was complicated by the fact that the design of
the missile itself was not complete and its dy-
namic fragmentation pattern was unknown. A
dynamic fragmentation pattern was assumed
from information supplied by the Services, but,
as shown in Section 1.5, the assumptions were
not strictly accurate. One very important com-
promise was made in the requirements for the
T-5 fuze from the ultimate Service needs. This
was in respect to the temperature range
throughout which the fuze could be used. Un-
impaired operation between —40 and +160 F
was desired, but because of the limitations of
the dry batteries which were to be used to power
the fuzes the low-temperature requirement was
waived. Actually, the relaxing of this require-
ment in the fuze did not impair the usefulness
of the complete weapon since the rocket itself
had low-temperature limitations not too dis-
similar from those of the fuze. In order to
reduce limitations due to possible deterioration
of the battery power supply during shipment
and storage, the design was made to allow final
assembly of the fuze in the field using freshly
tested batteries.

Experience gained in the development and
production of the T-5 fuze, combined with si-
multaneous investigations for improved power
supplies (Project SC-40), made possible much
expanded, more rigorous, and more specific re-
quirements for other radio proximity fuzes.
These included fuzes for the following: (1)
10,000-lb light-case [LC] bomb, (2) 4,000-lb
LC bomb, (3) 2,000-lb general purpose [GP]
bombs against both land- and water-borne tar-
gets, (4) 2,000-lb glider and controllable bombs,
(5) 1,000-lb GP bombs against water-borne

SECRET

OBJECTIVES AND MILITARY REQUIREMENTS

targets, (6) antiaircraft bombs for plane-to-
plane bombing, (7) fragmentation and anti-
materiel bombs of various sizes, and (8) large
chemical bombs of 500-, 1,000-, and 2,000-lb
sizes.

The military requirements for these bombs
were as follows:4*5

1. Adaptation to use in existing bombs, and
to fit and drop in existing bomb racks.

2. Strength enough to withstand handling
and shipping and, unarmed, drop safely on nor-
mal ground from 8,000 ft.

3. No deterioration from storage at temper-
atures from — 40 to +140 F.

4. A minimum of adjustment and assembly
in the field.

5. A design which minimizes the possibility
of triggering the fuze by enemy interference.

6. Suitability for day or night use.

7. Efficient operation at temperatures from
—40 to +140 F.

8. Efficient operation when released at any
indicated airspeed above 150 mph.

9. Efficient operation when released from al-
titudes up to 35,000 ft.

10. A minimum of 1,500 ft to arm.

11. Consideration in design toward evolving
a minimum number of fuze designs of suitable
performance necessary to meet the require-
ments of various sizes and types of bombs and
targets.

Burst heights were specified for only two of
the foregoing applications, the T-40 and T-43
fuzes for the 10,000- and 4,000-lb LC bombs.3
These heights were to be between 40 and 100 ft,
with the mean preferably near 50 ft. This was
believed to be the best height of operation for
enhanced blast effect from these large high-
charge bombs. The T-40 and T-43 were to be
tail fuzes. The sensitivities or operating heights
of the T-50 and T-51 fuzes intended for the
other applications were not defined. It was,
however, informally stated that, for antiper-
sonnel and antimateriel use, burst heights of
the order of 50 ft were desired. For the chem-
ical bombs, burst heights of the order of 500 ft
were believed best. Estimates in the former
case were based on theoretical computations of
fragmentation effect against shielded targets.11
The T-50 and T-51 fuzes were to be nose fuzes,

interchangeable with the M-103 contact fuze.

Following the development, production, and
service testing of the T-50 bomb fuzes, minor
changes, based on a fuller understanding of
their operational properties, were made in the
requirements for arming characteristics and
for burst heights. These changes led to models
T-89, T-90, T-91, and T-92, which are described
fully in Chapters 4 and 5.

Modifications were also requested in the T-50
type fuze to allow its use on Navy rockets,7 the
modified fuzes carrying the designations T-30
and T-2004 and differing from the T-50 mainly
in arming characteristics.

Experience gained in the development of the
T-50 and T-51 fuzes made it evident that the
physical size of radio proximity fuzes could be
reduced sufficiently to allow their use on trench-
mortar shells. Theoretical computations14 indi-
cated that a very appreciable gain in lethal
effect could be obtained by air-bursting such
shells. Accordingly, the Ordnance Department
requested the development of the T-132, T-171,
and T-172 fuzes10 for use on the 81-mm mortar
shells. According to military requirements,
these fuzes must :

1. Have a basic design also applicable to 105-
mm and 155-mm mortar ammunition.

2. Fit directly into the fuze cavity of stand-
ard 81-mm mortar ammunition.

3. Have sound ballistic design, minimizing
any deleterious effect on projectile drag and
stability as compared with fuzing with point
detonating fuzes.

4. When packaged, withstand rough han-
dling, shipping, storage over extended periods,
moisture, weather, and temperature cycles from
-40 to + 140 F.

5. When unpackaged, withstand loading op-
erations, moisture, weather, and temperature
cycles from —40 to +140 F for short periods,
and withstand rough handling expected under
service conditions incident to firing.

6. Be provided with a cap or cover to pre-
vent entry of mud or water into the air passage
after removal of the fuzed round or fuze from
its packaging, such can or cover to be removed
upon withdrawal of safety pin or pins.

7. Require a minimum of adjustment or
assembly in the field.

SECRET

INTRODUCTION

8. Function at or near optimum mean effec-
tive height on approach to ground over the
range of angles of fall encountered with these
projectiles.

9. Limit combined early bursts and duds to
15 per cent.

10. Not be readily affected by enemy jam-
ming or other interference.

11. Have a secondary element capable of
functioning on impact with minimum effect
and independently of the primary element.

12. Operate without detrimental interac-
tion, due to mutual interference, when fired at
random from weapons spaced closely together.

13. Have an interrupted detonator-explosive
train, safe against rough handling, dropping,
or crushing, until properly armed by removal
of safety pin or pins, acceleration of firing, and
a fixed air travel.

14. Have an arming delay mechanism which
will insure detonator safety up to 400 yd (ten-
tative estimated distance) from the mortar and
which will also delay fuze activation until flight
characteristics of the projectiles are sufficiently
stable to minimize early burst due to poor sta-
bility or action of the projectile and to permit
efficient fuze operation at the target.

15. Provide means for externally checking
the safe position of the arming mechanism.

16. Exhibit the above safety and operating
characteristics under the following conditions:
(1) temperature —40 to +160 F, (2) all
weather conditions, and (3) night or day.

The “mean effective height” referred to in
requirement (8), although not specified, was
understood to be of the order of 10 ft from
theoretical computations,14 but final specifica-
tions would have to await effect field trials.

It is to be noted that the requirements for
T-132, etc., are much more detailed and rigor-
ous than those for the T-5 fuze which had been
laid down three years previously. In particu-
lar, requirement (9) called for 85 per cent
proper functioning of fuzes, whereas 50 per
cent was allowed for the T-5.

One apparently innocuous requirement intro-
duced for security reasons applied to all bomb
and rocket fuzes developed by Division 4. This
was that all vacuum tubes used in the fuzes
were to fail at accelerations between 10,000

and 20,000<;.33 The purpose of the requirement
was to restrict the use of tubes suitable for
shell fuzes to that application, thereby reducing
the possibility that, through recovery of dud
fuzes by the enemy, shell fuzes would be copied
and used against our own air forces. As shown
in Chapter 3, this requirement introduced some
difficulties, because design considerations for
microphonic stability and for ruggedness are I
quite similar. Thus, in the course of developing
suitable antimicrophonic tubes for use in the
bomb and rocket fuzes, designs were developed
which were rejected because the tubes would
not fail at high accelerations. The requirement
was withdrawn in the fall of 1944 (when shell
fuzes were committed to battle under condi-
tions where they might be recovered by the
enemy) and thus did not apply to the mortar
shell fuzes developed by Division 4.

12 SELECTION OF THE DOPPLER-TYPE
RADIO PROXIMITY FUZE

The requirement that a fuze operate in the
vicinity of target may be fulfilled by making
the fuze sensitive to one of a variety of energy
forms: radio, optic, acoustic, magnetic, etc. A
comparison of the possibilities and limitations
of various energy-sensitive devices is given in
Volume 3, Chapter 2, of the Division 4 STR.
Here we are concerned only with radio meth-
ods.

Among the radio types there are two general
classes: active and passive. The active types
generate and radiate energy and are sensitive
to small amounts of energy after it is reflected
from a target. Passive-type fuzes are merely
sensitive to incident radio waves. In each of
these general classes there are further divisions
and subdivisions.

Active-type fuzes may operate by depend-
ence on interference between the original and
the reflected waves, or operation may depend
on the transit time for a pulse or train of waves
to travel from the fuze to the target and back
to the fuze again. Interference may occur in
several ways. If there is relative motion be-
tween the transmitter and the reflecting target,
the reflected waves when received at the fuze

SECRET

OPERATION AND COMPONENTS OF DOPPLER-TYPE FUZES

will differ in frequency from the transmitted
waves (doppler effect). Interference results in
a beat note equal to the difference in frequency.
On the other hand, if the transmitter is fre-
quency or phase modulated, interference with
the reflected waves produces a signal which is
a function primarily of the distance to the tar-
get. This principle is equivalent to that of the
well-known FM radio altimeter. Pulsed or in-
termittent circuits to determine time or dis-
tance to target operate on essentially the same
principles as the common forms of radar rang-
ing devices.

The simplest kind of passive proximity fuze
requires the target to be a source of energy.
Although this requirement can be satisfied for
antiaircraft fuzes of the acoustic or infrared
type, it would generally not hold for radio-
sensitive devices. Consequently, a passive-type
radio fuze would require auxiliary transmit-
ting equipment as part of the fire control.

In selecting an operating method for a radio
proximity fuze, probably the most important
consideration was simplicity. It was believed
that if the fuze was too complicated, it would
be impracticable on two grounds: (1) its vol-
ume would be too large to satisfy ballistic re-
quirements, and (2) it could not be manufac-
tured in sufficient quantities in time to be of
any value. Since fuzes are expendable devices,
to be used only once, an appreciably different
attitude toward production was required for
radio proximity fuzes than for other types of
radio equipment. Furthermore, a radio fuze is
a device on which no adjustment is possible
during its operation, hence reliability was a
requirement which could not be compromised
by the manufacturing problem. Thus, it ap-
peared imperative to keep the design of a radio
proximity fuze as simple as possible but still
fulfill the military requirements.

The simplest type of radio fuze is probably
the passive type, but, since auxiliary fire con-
trol equipment would be needed for its use, it
does not meet the general requirement for “a
minimum of special equipment and training”
for its operational use. Passive-type radio
fuzes were, however, seriously considered and
investigated until it was definitely established
that the transmitters required in active-type

fuzes could be built in large quantities and made
to operate reliably during the flight of the
missile.

Probably the simplest active-type radio fuze
is the doppler type, since the transmitter in
such a fuze requires no internal modulation or
control circuits other than an audio-frequency
amplifier. Furthermore, as is shown in detail
in Chapters 2 and 3, there are sufficient design
parameters available in doppler fuzes to adjust
the position of operation along the trajectory
of the missile approximately as desired.

All radio proximity fuzes developed by Divi-
sion 4 to the stage of adaptability to large-scale
production are based on the doppler principle.
Chosen initially because of its simplicity, the
basic method has proved adequate to meet the
major military requirements. More complicated
systems have been surveyed and tested briefly,
but none of these appeared simple enough to
reduce to a mass-production design in time to
be of value.

13 OPERATION AND PRINCIPAL

COMPONENTS OF DOPPLER-TYPE FUZES

The actuating signal in a doppler-type fuze
is produced by the interference with the trans-
mitter in the fuze of the reflected energy from
a target moving with respect to the fuze. The
frequency of the reflected energy differs from
the original by an amount (2v cos a)/X, where v
is the velocity of the fuze in a coordinate system
where the reflector is at rest, \ is the wave-
length of waves radiated by the fuze, and a is
the angle the velocity vector makes with the
line between the missile and target. The inter-
ference or combination of the two frequencies
produces a low-frequency signal equal to
(2v cos a) /l, which can be used to trigger an
electronic switch. Selective amplification of the
low-frequency signal is generally necessary.

It is shown in detail in Chapter 2 that the
concept of interference of the original and re-
flected waves is analytically equivalent to a load
variation on the transmitting oscillator. Hence,
an r-f circuit which responds to variations in
its loading will generate a target signal of fre-
quency ( 2v cos a) /l. This signal may be de-

SECRET

INTRODUCTION

tected in a separate mixing circuit, oscillator
diode [OD], or by a change in some parameter
of the oscillator circuit, such as grid voltage,
reaction grid detector [RGD] , or plate current,
power oscillating detector [POD]. Tl;e designa-
tions OD, RGD, and POD are further clarified
in Section 3.1.

The principal elements of a radio proximity
fuze are shown in block diagram form in Fig-
ure 1. The dashed lines between the oscillator
and detector indicate that the two functions
may be combined.

ANTE NNA

POWER SUPPLY ARMING

Figure 1. Block diagram showing principal
components of radio proximity fuze, doppler
type.

Operation of the fuze occurs when the output
signal from the amplifier reaches the required
amplitude to fire the thyratron. For a given
orientation of the fuze and target, the ampli-
tude of the target signal produced in the oscil-
lator-detector circuit is a function of the dis-
tance between the target and the fuze. Hence,
by proper settings for the gain of the amplifier
and the holding bias on the thyratron, the dis-
tance of operation may be controlled. Distance,
however, is not the only factor which requires
consideration. Orientation or aspect is very im-
portant, particularly against aircraft targets,
since operation should occur at that point on
the trajectory when the greatest number of
fragments will be directed toward the target.

For most missiles, the greatest number of
fragments are directed upon detonation ap-

proximately at right angles to the axis of the
missile. The dynamic fragmentation pattern
for an M-8 rocket is shown in Figure 2,b and its
essential features pertaining to fuze design are
typical of most missiles except that for higher-
velocity projectiles, the side lobes are inclined
forward toward the line of flight. The graph
shows the density of the fragments per unit
area of a sphere drawn about the missile as
a function of the angle between the direction
of the fragments and the axis of the missile.
The angle 0m represents the latitude angle along
which the greatest number of fragments are
directed. The three-dimensional pattern would
be that obtained by rotating the curve in Fig-
ure 2A about the flight axis. The static frag-
mentation pattern of a 500-lb GP bomb is
shown in Figure 2B. The dynamic pattern,
obtained by the vectorial addition of velocities
due to the bomb’s motion and due to the explo-
sion, would be tipped forward a few degrees.

For trajectories which would normally pass
by the target without intersecting it, there will
be optimum chance of damage if detonation of
the missile occurs when the target makes an
angle 0m with the missile. However, for trajec-
tories which would intersect the target, the
missile should come as close to the target as
possible before detonation. Hence, the basic re-
quirements for directional sensitivity of a
proximity fuze for antiaircraft use are (1) the
sensitivity should be a maximum in the direc-
tion corresponding to maximum lateral frag-
mentation density of the missile, and (2) the
sensitivity should be a minimum along the axis
of the missile. Directional sensitivity of this
type can be obtained by using the missile as an

b It was erroneously assumed during the development
of the T-5 fuze and in the absence of experimental data
that the latitude (dynamic case) of maximum fragment
density for the M-8 rocket would lie between 60 and 70
degrees. Actually the density of lethal fragments in
this direction is greater than shown in Figure 2A be-
cause the contribution of the relatively low-velocity
fragments from the rocket body are not shown in the
figure. For high-velocity missiles, such as antiaircraft
shells, the component of velocity due to the shell’s for-
ward motion gives a very appreciable forward tilt to
the dynamic fragmentation pattern. Also, in the case
of higher-velocity aircraft rockets [HVAR] such as the
5-in. HVAR, the latitude of maximum fragmentation
density is about 66 degrees. Fuzes for this rocket (T-30)
were developed later in World War II.

SECRET

OPERATION AND COMPONENTS OF DOPPLER-TYPE FUZES

antenna with the axis of the missile corre-
sponding to the axis of the antenna. With the
fuze in the forward end of the missile, such an-
tennas are end-fed by means of a small elec-
trode or cap on the nose of the fuze. Additional
control over the sensitivity pattern of the fuze
is possible by means of the amplifier gain char-
acteristic. As pointed out previously, the fre-

For use against surface targets, proximity
fuzes are designed for an optimum height of
burst, depending on the nature of the target
and the properties of the missile. When frag-
mentation bombs are air burst, the possible
damage to shielded targets is substantially in-
creased. Figure 3 shows a cross-sectional view
of a typical shielded target: a man in a fox-

Figure 2. Fragmentation patterns of missiles.

The amplitude represents the relative density of the fragment as a function of the latitude angle around the axis of the missile.
Figure 2A is for the M-8, 4.5-in. rocket and shows the dynamic pattern; i.e., directional allowance has been made for the effect
of the velocity of the rocket. The graph, which is based on data in reference 19, does not include the contributions of fragments
from the rocket motor. The latter are large, slow-moving, relatively few in number, and add to the pattern shown in the region
between 45° and 90°. Figure 2B is a static pattern for the M-43, 500-lb GP bomb and is based on data in reference 11. The effect
of bomb velocity on fragment direction is very slight (due to the relatively low velocity of the bomb) and would shift the maximum
of the pattern forward of the order of 5°.

quency of the target signal is {2v cos a) /l. The
angle a varies rapidly as the missile passes the
target, and if maximum gain occurs when
a = 6m there will be greater likelihood that the
missile will be detonated at the proper point on
its trajectory. More detailed discussion of these
features is given in Sections 2.8 and 2.11 and
in Sections 3.2 and 3.5.

hole. The man is shielded from fragments from
any bomb detonating either side of the hole and
below the dashed lines. The angles <f>R and <f>L,
which the lines make with the horizontal, are
called the shielding angles for the respective
directions. It is thus seen that, as the <j> s
increase, higher burst heights will be neces-
sary to expose the targets. An upper limit on

SECRET

INTRODUCTION

burst height is set by the lethal range of the
bomb fragments since these fragments lose
velocity rapidly as they travel from the point
of explosion. Hence, the height of an air burst
should be great enough to expose an appreci-
able number of targets but not so high that the
fragments will be impotent when they strike
the targets.

Most computations and evaluation tests for
the optimum height of air burst for bombs
have been on the basis of a 10° shielding or
safety angle.c The optimum height varies only

Figure 3. Sectional view of a soldier in fox-
hole, typical shielded target. Soldier is protected
from fragments from explosions below dashed
lines. Angle these lines make with horizontal are
called shielding, or safety, angles.

slightly for the various striking angles and
velocities with which bombs may approach the
ground. Hence, it is desirable to design a fuze
for ground-approach use which will give essen-
tially constant burst heights for the various
approach conditions.

An approach to this requirement is to have
maximum radio sensitivity along the axis of
the bomb, with essentially constant sensitivity

c See references 11, 12, and 13 for theoretical values
and reference 16 for effect field tests. It should be
pointed out that the size of the elementary target is a
primary consideration in the computation of optimum
burst heights. From this point of view, overhitting on
targets of finite size is decreased as the burst height
increases. Thus, an optimum burst height is determined
by the lethal range of fragments on the one hand and a
height where overhitting becomes excessive on the other.

to about 45 degrees on either side of the axis.
(For most release conditions used operation-
ally, bombs strike the ground with an angle to
the vertical of less than 45 degrees.) A short
dipole antenna mounted transversely to the
bomb’s axis and on the nose of the bomb essen-
tially meets this requirement. In addition, it is
necessary to design the amplifier of the fuze to
give constant amplification for the range of
doppler frequencies which might be encoun-
tered because of various approach velocities.

On the other hand, it was found that fairly
good ground-approach performance could be
obtained with fuzes with axial antennas by de-
signing the amplifiers to compensate for the
appreciable decrease in radiation sensitivity in
the forward direction. For example, steep
angles of approach in general mean high ap-
proach velocities with higher doppler frequen-
cies. Thus, a loss in radiation sensitivity with
steep approach can be compensated by an in-
crease in amplifier gain for the higher doppler
frequency. Details of such design are given in
Section 2.2.

A miniature triode is used for the oscillator
in the fuze and a pentode for the amplifier.
When a separate detector is used, a tiny diode
provides the required rectification. A miniature
thyratron serves as the triggering agent and
a specially developed electric detonator initiates
the explosive action. Details concerning the de-
sign of these elements are presented in the vari-
ous sections of Chapter 3.

Energy for powering the electronic circuits
is obtained in the later fuze models from a
small electric generator. This is driven by a
windmill in the airstream of the missile. A rec-
tifier network and voltage regulator are essen-
tial parts of the power supply. Design details
of the generator power supply, as well as
earlier battery power supplies, are given in
Section 3.4.

The arming and safety features of the radio
proximity fuzes are closely tied in with the
power supply. This is a natural procedure since
an electronic device is inoperative until electric
energy is supplied. Arming a radio proximity
fuze (generator type) consists of the following
operations: (1) either (a) removal of an arm-
ing wire which frees the windmill, allowing it

PRODUCTION OF PROXIMITY FUZES

to turn in the airstream (bomb fuzes) or (b)
actuation of a setback device freeing the drive
shaft of the generator and allowing it to turn
(rocket and mortar-shell fuzes), (2) operation
of the generator to supply energy to the fuze
circuits, (3) connection of the electric detona-
tor into the circuit after a predetermined num-
ber of turns of the vane corresponding to a
certain air travel, and (4) removing a mechani-
cal barrier between the detonator and booster
prior to which explosion of the detonator would
not explode the booster. Generally, operations
(3) and (4) occur simultaneously by motion
of the same device.

Additional safety is provided by the fact that
unless the generator of the fuze is turning rap-
idly the fuze is completely inoperative. A mini-
mum airspeed of approximately 100 mph is
required to start the generator turning. Details
of the arming system are given in Chapter 4.

14 PRODUCTION OF PROXIMITY FUZES

The course of the development of radio prox-
imity fuzes for fin-stabilized missiles and the
actual nature of the devices placed in produc-
tion for Service use were influenced by many
factors other than fundamental technical con-
siderations. Time and expediency had a major
influence on all designs. In order to have fuzes
available for use as soon as possible, tooling for
large production was frequently started before
development was complete. This meant that
when further development indicated certain de-
sign changes to be imperative or desirable the
extent of the changes which were made was
controlled by the degree of the changes required
in tooling or by the amount of time which
would be lost by making the changes. Further-
more, no components could be included in the
design which would take too long to acquire in
the necessary quantity nor could production
techniques be considered which were over-
elaborate and time consuming.

Specific Service requirements varied as the
course of World War II changed, and, because
of the pressing demand for speed, fuze designs
for the new requirements made much more use

of the tools and techniques employed in preced-
ing models than if production had started out
fresh. For example, the greatest urgency early
in World War II was for antiaircraft weapons,
and stress was placed on fuzes for both bombs
and rockets for this purpose. When the Allies
acquired undisputed air superiority, the major
proximity fuze requirements were shifted to
the ground-approach operation. Thus, the T-50
type bomb fuze, which employs the axial radio
antenna, ideal for antiaircraft use and initially
designed for that purpose, was adapted to
ground-approach use. The T-51 fuze, which em-
ploys the transverse antenna specifically devel-
oped for ground-approach use, was used much
less extensively for this application because its
initial lower priority made it available later in
World War II.

More detailed information relating to the
sequential development of radio proximity
fuzes is given in the history of Division 4. The
subject is mentioned here only to emphasize
that the technical phases of the development
were not always controlled by straightforward
engineering design considerations.

After the operation of a fuze design was
found satisfactory by laboratory and field tests,
it was necessary to determine its practicability
for mass production. Pilot construction lines
were used for this purpose, and it was the
policy of Division 4 to require the construction
of about 10,000 pilot-line fuzes with suitable
performance characteristics before releasing a
design to the Armed Services. Usually the tools
developed for the pilot-line work were used also
for final production. Large-scale procurement
was handled by the Services, but Division 4
participated in many phases of it, largely in an
advisory capacity. The various technical aspects
involved in the production of radio proximity
fuzes are presented in Chapter 6.

The radio proximity fuzes developed by Divi-
sion 4 to the stage of large-scale production are
as follows. More detailed information concern-
ing the characteristics of these fuzes is given in
Chapter 5.

M-8 Rocket Fuzes

1. T-5, an antiaircraft battery-powered fuze
for the 4.5-in. M-8 rocket. This fuze is shown in

SECRET

INTRODUCTION

Figure 4. Approximately 370,000 were pro-
cured by the Army.

2. T-6, a ground-approach fuze, for use as an
artillery weapon on the 4.5-in. M-8 rocket. This
fuze is a variation of the T-5, having a longer

Figure 4. Radio proximity fuzes for rockets.
These are from left to right: (1) T-5 fuze for
4.5 in. M-8 rocket for air-to-air use (T-6 ground-
to-ground fuze is identical in appearance to
T-5) ; (2) T-2004 fuze for 5-in. AR rocket for
air-to-ground use; and T-2005 multiple-purpose
fuze.

arming time, about 6 sec compared to 1.0 sec,
and no SD element. It is identical in exterior
appearance to the T-5. Approximately 300,000
of the T-5 fuzes were converted after comple-
tion to T-6 fuzes.

3. T-12, a generator-powered fuze for use on
the M-8 rocket. This fuze was not placed in
large production primarily because of curtail-
ment in requirements for the M-8 rocket.

Bomb Fuzes

1. T-50-E1, a generator-powered ground-
approach fuze for use primarily on the 260-lb
M-81 fragmentation bomb, the 100-lb M-30 GP
bomb, and the 2,000-lb M-66 GP bomb. This
fuze, which uses the bomb as a radio antenna,
was planned for air-to-air use when develop-
ment started but was changed to ground-
approach application before development was
completed. Its radio transmitter operates in the
Brown frequency band. This fuze was set to

arm after 3,600 ft of air travel. It is shown in
Figure 5.

2. T-50-E4 is similar to the T-50-E1 fuze ex-
cept that its transmitter operates in a different
frequency band (White band), giving optimum
performance on the 500-lb M-64 and the 1,000-
lb M-65 GP bombs. Approximately 130,000
T-50-E4 and T-90 fuzes were procured by the
Army.

3. T-89, an improved T-50-E1 type fuze, giv-
ing more uniform burst heights. It also differs
from T-50-E1 in that arming setting can be
checked more readily in the field. Approxi-
mately 140,000 T-50-E1 and T-89 fuzes were
procured by the Services. This fuze is similar
in appearance to the T-91 fuze, shown in Fig-
ure 5.

4. T-91 (later designation, M-168), a varia-
tion of the T-89, developed specifically to meet a
naval requirement of higher burst heights than
the T-89 for low-altitude bombing. This fuze
is set to arm after 2,000 ft of air travel. Ap-
proximately 120,000 T-91 fuzes were produced.

5. T-92, a variation of the T-90 developed to
meet the same performance requirement as the

Figure 5. Radio proximity fuzes for bombs.
These are, from left to right: (1) T-50-E1 fuze
for air-to-ground use on M-30 and M-81 bombs;

(2) T-91 fuze, a later and improved version of
T-50-E1, for use on M-30, M-81 and M-64 bombs;

(3) T-51 fuze, air-to-ground use, for use on all
bombs of 100-lb size or larger; and (4) T-82
fuze for use paralleling T-51.

T-91 of higher burst heights in low-altitude
bombing. It is similar in appearance to the T-91
fuze. Approximately 70,000 were produced.

6. T-51 (later designation, M-166), a gen-
erator-powered bomb fuze with a transverse
antenna for ground-approach use on all GP,
fragmentation, and blast bombs of 100-lb size

GENERAL EFFECTIVENESS OF PROXIMITY FUZES

or larger. Burst heights with the T-51 are gen-
erally higher than with T-50 type fuzes. This
fuze was set to arm after 3,600 ft of air travel.
Approximately 350,000 were procured by the
Services.

7. T-82, a generator-powered bomb fuze
with transverse antenna of somewhat different
physical dimensions than the T-51. It was de-
veloped for the same general purpose as the
T-51, but, when success of the latter was as-
sured, further development of the T-82 was
turned over to the Army.8 It had not reached
the production stage at the time of the transfer.

Later Rocket Fuzes

1. T-30 (Navy designation, Mk-171), a gen-
erator-powered rocket fuze for air-to-air use,
particularly on the Navy’s HVAR and the 5-in.
aircraft rocket [AR]. This fuze is physically
very similar to the T-91 bomb fuze and only
slightly different electrically. Its arming sys-
tem is different in that the acceleration of the
rocket is essential to its operation. This fuze
had just reached a production rate of 10,000
per month at the end of World War II.

2. T-2004 (Navy designation, Mk-172), a
generator-powered rocket fuze for ground-
approach use. It is similar to the T-30, but is
somewhat less sensitive and has a longer arm-
ing time. Approximately 110,000 were pro-
cured by the Services. A photograph is shown
in Figure 4.

3. T-2005, a miniature generator-powered
rocket fuze for either antiaircraft or ground-
approach use (by a change-over switch). It is
similar electrically to the T-30 and T-2004. De-
velopment of this fuze was initiated by Divi-
sion 4 but turned over to the Army for further
work before the point of large-scale production
was reached. A photograph of the fuze is shown
in Figure 4.

Trench-Mortar Fuzes

1. T-132, a generator-powered ground-ap-
proach fuze for use on the 81-mm trench-mor-
tar shell. This fuze, shown in Figure 6 along
with the T-171 and T-172, uses the body of the
shell as an antenna. It also incorporates a novel
production technique, i.e., printed or stenciled
electric circuits. Tools were being set up for a

production rate of approximately 100,000 per
month when World War II ended.

2. T-171, a generator-powered ground-ap-
proach mortar-shell fuze, similar to the T-132

Figure 6. Radio proximity fuzes for trench
mortar shells. These are, from left to right:

(1) T-132 fuze using electric circuits “printed”
on ceramic plates; (2) T-171 fuze, electrically
similar to T-132 but with standard electrical re-
sistor and condensers; and (3) T-172 fuze with
loop antenna.

except that it employs the more standard cir-
cuit-assembly techniques. Tools were being set
up for production rate of about 125,000 per
month when World War II ended.

3. T-172, a generator-powered ground-ap-
proach mortar-shell fuze with a loop antenna.
This antenna has essentially the same direc-
tional properties as the transverse antenna of
the T-51 bomb fuze. Tools were being set up for
a production rate of about 250,000 per month.

Development of the T-40 and T-43 bomb tail
fuses (referred to in Section 1.2) for the 4,000-
and 10,000-lb blast bombs was not completed
because the T-51 nose fuze appeared to be ade-
quate to meet all the requirements. As shown
in Chapter 9, tests of the T-51 fuzes (with
minor modifications) on M-56 (4,000-lb) bombs
gave excellent performance. No 10,000-lb
bombs were made available for field tests.

15 GENERAL EFFECTIVENESS OF
PROXIMITY FUZES

Although the final answer on the effective-
ness of a new military weapon is supplied by

SECRET

INTRODUCTION

its performance in battle, the best quantitative
measure of relative effectiveness under con-
trolled conditions can be obtained from care-
fully planned field trials. A number of evalua-
tion tests have been carried out on radio
proximity fuzes. These can be grouped into the
following categories.

1. Evaluation of conformance to require-
ments.

2. Evaluation as a weapon:

a. Antiaircraft use (fragmentation ef-
fect) .

b. Air burst (ground approach on frag-
mentation bombs and rockets).

c. Air burst on blast bombs.

d. Air burst on chemical bombs.

e. Air burst on fire bombs.

Most of the tests conducted by Division 4
other than strictly developmental tests were in
the first category above. The Services also car-
ried out extensive tests in the first category but
generally after the fuzes were in production.

Tests and evaluation studies in category 2
above were usually carried out by the Services
or by other NDRC divisions and therefore are
not properly within the scope of this volume.
The results, however, are of interest in giving
a more complete picture of the evaluation of
radio proximity fuzes and accordingly will be
referred to briefly. Such evaluations, of course,
depend primarily on the properties of the mis-
sile which carry the fuzes and in no cases were
the missiles designed for proximity operation.
Now that proximity fuzes have been established
as practicable devices, certain missiles, such as
fragmentation bombs for air-burst use should
be redesigned to increase greatly their effec-
tiveness as weapons.

Typical missiles equipped with proximity
fuzes are shown in Figure 7.

Evaluation of Conformance to
Requirements

Detailed evaluations of the conformance of
the fuzes to the military requirements are pre-
sented in Chapters 5 and 9. In this section, a
brief abstract is given of the most important
results for production fuzes. Generally, the re-

liability of the radio proximity fuzes for bombs
and rockets was about 85 per cent, that is, 85
per cent of the fuzes would be expected to func-
tion on the target as required. Of the remainder
about 10 per cent could be expected to function
before reaching the target (random bursts)
and 5 per cent not to function at all. The 10 per
cent or so random functions were distributed
along the trajectory between the end of the
arming period and the target. In many thou-
sands of tests, no fuze functions were observed
before the end of the arming period.

Figure 7. Radio proximity fuzes on typical
missiles. These are, from bottom to top: (1)
T-132 fuze on 81-mm M-56 mortar shell; (2)
T-91 fuze on M-81-A 260-lf fragbentation bomb;

(3) T-51 fuze on the M-64, 500-lb general pur-
pose bomb; and (4) T-2004 fuze on 5-in. HVAR
rocket.

General reliability and proximity sensitivity
(function heights) for the various production
models follow.

1. T-5 Fuze. Acceptance tests on over 4,000
T-5 fuzes against a mock airplane target
showed the following results:

a. 81 per cent proper functions in the
vicinity of the target.

b. 2 per cent functions just beyond the
target.

c. 13 per cent early functions between
arming and the target.

d. 4 per cent duds.

The time of flight in normal acceptance tests
(see Chapter 8) was inadequate to allow test-

GENERAL EFFECTIVENESS OF PROXIMITY FUZES

in g of the SD feature. Separate tests on the SD
showed it to be 96 per cent reliable at an aver-
age time of 8.5 sec after firing. Ninety per cent
of the SD functions were between 6.5 and 11
sec. These figures refer to the mechanical SD
(see Chapter 4) used in later models. An elec-
tric SD used in earlier models (see Section 3.3)
was less reliable.

The vicinity of the target was defined as
within a 60-ft impact parameter of an 0.8-scale
target of a medium bomber. For more detailed
discussion of a proper definition of “vicinity of
the target” see Section I.5.2.1921

2. T-6 Fuze. The percentage of proper func-
tions for the T-6 ground-approach fuze depends
on the time of flight of the rocket, the number
of random functions increasing with the longer
trajectories. For maximum range, tests on over
1,500 rounds indicated the following perform-
ance.

a. 80 per cent proper functions.

b. 16 per cent random functions.

c. 4 per cent duds.

Proper functions were defined as operation
between 6 and 100 ft above ground.

3. T-50-E1 and T-89 Fuzes. Acceptance tests
on 100 lots (lots averaged about 1,000 fuzes
and field tests were made on about 18 fuzes
from each lot) of T-50-E1 and T-89 bomb fuzes
showed

a. 83 per cent proper functions.

b. 13 per cent random functions.

c. 4 per cent duds.

Proper functions for ring-type bomb fuzes
(axial antennas) were defined as between 6 and
100 ft over a water target. The average burst
height was 33 ft.

4. T-91 Fuzes. The first lots of T-91 bomb
fuzes were about the same quality as the T-50-
E1 fuzes. However, later lots (T-91-E1 using
the RGD circuit, see Section 3.1) showed the
following average for 27 lots.

a. 92 per cent proper functions.

b. 7 per cent random functions.

c. 1 per cent duds.

The average height of function was 60 ft over
a water target.

5. T-50-E4 and T-90 Fuzes. Tests on 130 lots
of T-50-E4 and T-90 bombs showed

a. 78 per cent proper functions.

b. 19 per cent random functions.

c. 3 per cent duds.

The average height of function was 40 ft.

6. T-92 Fuzes. Tests on 50 lots of T-92 bomb
fuzes showed

a. 58 per cent proper functions.

b. 34 per cent random functions.

c. 8 per cent duds.

The average height of function was 34 ft.

The inferior performance of T-92 fuzes was
due to unusual dependence of the fuze on the
electric properties of the test missile, the M-64
bomb. It was found that, on bombs which had
been carefully prepared to reduce variable con-
tact between the fin and the bomb body, scores
equal to those with other fuzes were obtained.
When it was definitely established that the poor
performance of the T-92 was due to this cause
and consequently could not be improved by
more rigorous production control, further pro-
curement was terminated. It had meanwhile
been shown that the T-51 and T-91 fuzes, which
had become available, would fulfill the applica-
tions for which the T-92 was intended.

7. T-51 Fuzes (M-166). Field tests on 230
lots of T-51 bomb fuzes showed

a. 91 per cent proper functions.

b. 9 per cent random functions.

c. 1 per cent duds.

The average height of function over the water
target was 110 ft. The proper function range
included heights up to 200 ft for bar-type fuzes.

8. T-2004 Fuzes. Field tests on 75 lots of
T-2004 rocket fuzes showed

a. 94 per cent proper functions.

b. 3 per cent random functions.

c. 3 per cent duds.

The average height of the proper functions was
30 ft.

1,5,2 Evaluation as a Weapon

Antiaircraft Use

A careful analysis of the T-5 fuze on the M-8
rocket as an antiaircraft weapon was made by
the Applied Mathematics Panel [AMP].19'21
The study was based on the experimental per-
formance of the fuze against a mock aircraft
target, fragmentation data of the rocket, dis-

INTRODUCTION

persion data on the rocket when fired from an
airplane, and vulnerability of a twin-engine
enemy aircraft, in particular the JU-88, to
fragmentation damage.

Conclusions of these studies were as follows :

1. When fired from 1,000 yd directly astern
with a standard deviation in firing error of
about 50 ft (17 mils), a single round has one
chance in * 10 of preventing a twin-engined
bomber from returning to base provided it
cannot return to base on one engine.

2. If return to base on one engine is pos-
sible, there is one chance in 16 that a single
round will prevent its return.

3. If a delay of about 50 ft were incorporated
in the fuze, to bring the vulnerable part of the
target in a region of greater fragmentation
density, the above probabilities would be in-
creased to 1 in 4 and 1 in 6.

The greater effectiveness of the weapon with
the delay was due to the fact (as shown in Fig-
ure 2) that the latitude of greatest fragmenta-
tion density of the rocket was approximately at
right angles to the axis of the rocket, whereas
the fuze, as shown in Chapters 2 and 3, had
been designed from an assumed latitude of
maximum density of about 70 degrees. A delay
of the amount recommended in the AMP report
would have brought the target in the region of
maximum fragment density. Such a delay could
have been incorporated readily in the fuze had
the tactical demand for this weapon in 1944
been as high as it was in 1942. However, there
appeared to be little likelihood that M-8 rockets
would be used as air-to-air weapons, so the
fuzes were not modified.

The probability of obtaining a crippling di-
rect hit by an M-8 fired under the same condi-
tions is about 1 in 100.

Limited tests and evaluations were made of
the 5-in. AR and HVAR equipped with T-30
fuzes as antiaircraft weapons. At the Naval
Ordnance Test Station at Inyokern, California,
some 70 rounds were fired from a fighter air-
plane at a radio-controlled plane in flight.18 At
about 400-yd range, over 55 per cent of the
rounds functioned on the target. Eight high-
explosive [HE] loaded rounds were fired, four
of which functioned on the target, and three of
the four destroyed the targets. Presumably,

most of the rounds which did not function on
the target were beyond the range of action of
the fuzes.

The Applied Mathematics Panel made an in-
formal study of the effectiveness of AR and
HVAR equipped with proximity fuzes.22 For
these rockets it was found, presumably because
of their higher velocities, that the optimum
burst surfaces were inclined forward from the
equatorial plane of the rocket and not at right
angles to it, as was the case for the M-8 rocket.
No experimentally determined burst surface
patterns were obtained for T-30 fuzes but,
assuming the same burst pattern as for T-5
fuzes, the effectiveness was nearly optimum.
For example, the probability of destroying an
aircraft with an HVAR with a firing error of
25 mils at 1,000-yd range was 0.4, and with 15
ft optimum delay was 0.63. Further details are
in the AMP report.

Air Burst for Ground Targets

The Army Air Forces [AAF] carried out ex-
tensive evaluations of the effectiveness of air-
burst bombs against shielded targets using
T-50 and T-51 fuzes on M-81 (260-lb fragmen-
tation) and M-64 (500-lb GP) bombs. Bombs
were dropped on a large effect field covered
with target boards 2x6 in. in trenches 1 ft deep.
For equivalent airplane loads of properly func-
tioning bombs dropped on 12-in. deep trench
targets, conclusions from the AAF report
are :16

1. Air-burst 260-lb M-81 fragmentation

bombs and 500-lb M-64 GP bombs produce
about 10 times as many casualties as contact-
burst 20-lb M-41 fragmentation bombs when
trenches are 15 ft apart. (A casualty is defined
as one or more hits per square foot, capable of
perforating % in. of plywood.)

2. Optimum height of burst for maximurfi
casualty effectiveness is between 20 and 50 ft,
with only slight variation through this range.

The British carried out similar appraisals,
using T-50 fuzes on M-64 bombs.26 There are
several differences in details of the tests, par-
ticularly in the matter of evaluating the effec-
tiveness of surface-burst bombs. The British
Ordnance Board made an appreciable allow-
ance for the blast effect of both the contact-

GENERAL EFFECTIVENESS OF PROXIMITY FUZES

fuzed bombs and variable-time [VT] fuzed
bombs and arrived at a superiority factor of
4 to 1 for the latter against shielded or en-
trenched targets.

The AAF also evaluated the M-8 as an air-
to-ground weapon with both VT (T-5) and
contact fuzing.17 The summary report con-
cluded that the weapon was relatively ineffec-
tive against shielded surface targets, although
the casualties per round with VT fuzing were
about five times as high as with contact fuzing.

Air Burst for Blast Bombs

Studies by Division 2, NDRC,23 and by the
British25 demonstrated that when large blast
bombs are air burst at about 50 to 100 ft above
ground, the area of demolition could be in-
creased from 50 to 100 per cent. No full-scale
tests were carried out to verify these conclu-
sions, but it was established that the T-51 fuze
could be used on both the 4,000-lb (M-56)
American bomb (Chapter 9) and the 4,000-lb
British bomb27 to give air bursts at the proper
altitudes.

In cooperative tests by the Army, Division 4
and Division 2, NDRC,24’34 it was shown that
air-burst bombs could be used in mine-field
clearance. The advantages were primarily in
increased reliability of clearance and absence
of cratering. However, the use of air-burst
bombs for this purpose does not markedly re-
duce the number of bombs required to clear an
area.

Air Burst for Chemical Bombs

A number of evaluations were made to deter-
mine the effectiveness of air bursts on chemical
bombs. In a carefully planned effect field test
using T-51 and T-82 fuzes on 500-lb LC bombs,
the British showed the areas of contamination
with a mustard-type gas were 4 to 5 times
greater than when the bombs were used with
contact fuzes.30 The increase was due to a more
uniform distribution of the vesicant and avoid-
ance of loss of material in craters.

The Chemical Warfare Service and the Brit-
ish cooperated in an extensive series of tests at
Panama in simulated jungle warfare. A T-51
fuze with reduced sensitivity effectively pro-
duced air bursts of chemical bombs below tree-

top canopies with efficient distribution of chem-
ical materials.28’ 29

Air Burst for Fire Bombs

The Army Air Forces evaluated the effective-
ness of T-51 fuzes on fire bombs and found that
for high-altitude bombing the distribution of
incendiary material was appreciably improved.
In this application, the gain due to an air burst
was due to the elimination of loss of material
in craters.31

1 5,3 Operational Use of Proximity Fuzes

Proximity fuzes for bombs and rockets saw
very limited operational use, primarily because
they were introduced into action very late in
World War II. Some of the factors which im-
peded their initial operational use are dis-
cussed in the history of Division 4. Other fac-
tors, as well as a full summary of their use in
World War II, are given in a memorandum by
a member of the VT Fuze Detachment of the
Ordnance Department.32 Some excerpts from
the latter reference are given in Chapter 9.

Altogether, approximately 20,000 fuzes, pri-
marily bomb fuzes, were used in action by the
Army and the Navy in the Pacific, and in the
European and Mediterranean Theatres of Op-
eration [ETO] and [MTO]. In the last few
weeks of the war in the Pacific, approximately
one-third of all bomb fuzes used by carrier-
based aircraft were proximity fuzes. The main
targets were antiaircraft gun emplacements
and airfields.

No thoroughgoing analysis of the effective-
ness of the fuzes operationally was possible,
although the general reaction was very favor-
able. Since the fuzes were used in all theaters
so late in World War II, the major uses were
of a trial or introductory nature. In all cases,
these trial uses were followed by urgent re-
quests for more fuzes, which usually, and par-
ticularly in ETO and MTO, did not arrive until
after World War II was over. All initial uses
were in 1945, in February in the Pacific and in
March in ETO and MTO. Reports concerning
the effectiveness of the fuzes against gun em-
placement targets generally stated that anti-
aircraft fire was either stopped or greatly re-

SECRET

INTRODUCTION

duced after the air-burst bombs exploded.

Although relatively little or no quantitative
data as to the effectiveness of the fuzes was
secured, their use was extensive enough to es-
tablish their practicability as service items of
ordnance equipment. Relatively little difficulty
was experienced in the handling and use of the
fuzes and none of these was serious or insur-
mountable. Hence, with the effectiveness of
proximity fuzes well established by effect field

studies and operational practicability estab-
lished by combat use, proximity fuzes appear
assured of a permanent and increasingly im-
portant position in modern ordnance. The tech-
nical information presented in the succeeding
chapters of this volume not only serves to pro-
vide a full understanding of the properties of
the fuzes which were developed and produced,
but it also provides a firm and logical basis for
future development.

SECRET

Chapter 2

THE RADIATION INTERACTION SYSTEM

21 INTRODUCTION

rj 1 he PRECEDING chapter has explained what
1_ a proximity fuze is and has shown what the
fuze must do by stating the military character-
istics required for such a device. The basis
upon which the radio reflection principle was
selected as most suitable for a proximity fuze
has been discussed, and some of the reasons for
using the doppler principle have been pre-
sented. We are now in a position to explain the
working principles of the device and its engi-
neering design.

In discussing the working principles we are
concerned with two essentially independent
sets of phenomena: (1) those external to the
fuze mechanism, i.e., the emission and recep-
tion of radiation and its interaction with the
target; and (2) those within the fuze itself,
i.e., internal circuit behavior.

The present chapter deals with the first
group, external phenomena, which we call the
radiation interaction system. To facilitate dis-
cussion, an arbitrary dividing line is drawn at
the point where the internal fuze circuit is
electrically connected to the fuze antenna. As
will be seen, it is possible to describe the ex-
ternal phenomena so that their effect can be
expressed as an appropriate variation of im-
pedance at these antenna terminals. When the
relation between the radiation interaction with
the target and the variation of antenna im-
pedance has been determined, the problem be-
comes one of constructing a practical circuit
which will respond properly to the changes
seen at its terminals.

It should not be inferred from this division
of phenomena, for the purposes of discussion,
that antenna design and circuit design are en-
tirely independent. Each must be designed with
due regard to the other, and both designs are
dictated by such practical considerations as
physical limitations of components and tactical
utility. In fact it will become evident as the dis-
cussion proceeds that the working principles of

a By R. D. Huntoon and P. R. Karr.

the fuze are quite simple and that the real dif-
ficulty in making a practical proximity fuze lies
in reaching an adequate compromise between
a host of closely interrelated factors. The co-
ordination of these various factors is treated
in Section 3.5.

Many of the phenomena treated in this chap-
ter are shown to be negligible or unimportant
for the type of doppler fuzes of immediate in-
terest. The phenomena may, however, have
appreciable importance for fuzes of other types
or for more extensive applications of the pres-
ent fuzes. For these reasons, the basic theory
has been treated in appreciable detail by de-
veloping considerable material found in ad-
vanced textbooks on radiation and circuit
theory. This approach should enable new in-
vestigators in the field of proximity fuzes to
familiarize themselves with the fundamental
principles involved with a minimum of re-
course to the technical literature.

22 SPECIFICATION OF PROBLEM IN
TERMS OF ANTENNA IMPEDANCE

The fuze detects the presence of an obstacle
in its radiation field by means of returning
radiation reflected by the obstacle. The physical
situation is thus characterized by an outgoing
wave with a frequency determined by the fuze
transmitter and a returning wave of much
smaller amplitude, whose frequency may be
different as a result of relative motion of fuze
and reflector. In all the discussion which fol-
lows, it will be assumed that the reflecting ob-
stacles are linear reflectors, by which we mean
that the strength of the reflected field is propor-
tional to the strength of the incident field. It is
shown in this section that the returning wave
differs in frequency from the outgoing wave by
an amount which can be calculated by the appli-
cation of the doppler principle, and that under
certain conditions, which hold for present fuze
designs, this combination of outgoing and re-
flected wave is exactly equivalent to a change

SECRET

THE RADIATION INTERACTION SYSTEM

in antenna impedance. The usefulness and limi-
tations of this concept are discussed. f

2-21 Reflected Wave or Doppler
Frequency Concept

Consider a radiating system R which radiates
a carrier of frequency /. Its field in any direc-
tion x will be of the form

E = (i)

Let the radiation be received in a system R'

moving with a velocity v in the direction —x,
i.e., toward the system R. In this moving system
of reference the field will be of the form

E' = A'ei2rf' [t' ~ (x'/c)l. (2)

The phase of the wave is relativistically invari-
ant, so that

7) »)

Now V and x' are related to t and x by the

Lorentz transformation. Applying this gives

when it is remembered that v is —dx/dt.

At the present time relative velocity of fuze
and target never exceeds 5,000 fps so that v/c
is of the order of 5 X 10-6. Equation (4) can be
rewritten

(5)

r =f(1+k + lcft) = f + i(1 +
which is close enough to

r = / + J (6)

which is recognized to be the normal doppler
frequency shift. Thus the frequency received at
the target is given by equation (6) above. The
target reradiates, reflects, on this frequency,
and a second application of the above argument
leads to

f" = / + p (7)

where /" is the frequency of the reflected wave
as seen at the fuze. For current fuze designs,
the doppler or difference frequency 2v/l is of

the order of a few hundred cycles per second
out of a carrier frequency of the order of 100
me and the error introduced by neglecting rela-
tivistic effects is of the order of 10-4 c.

2 2 2 Reflection Equivalent to Change of
Antenna Impedance

The two-wave picture outlined above can be
converted to the equivalent impedance picture
quite simply. First, assume the system R' to be
at rest, so that f — f' =/". Then the field of the
system R, equation (1), can be written as

E = KIej2v[ft - - (8)

where the dependence of E upon I, the antenna
current, is shown explicitly. This field is re-
flected from the target at distance x with a loss
in amplitude and a phase shift 5 and returns as
reflected field Er, given by

Er = BKIeW* ~ (2*/x) + 5]. (9)

The constant B represents the loss at reflection
and represents also the initial assumption that
reflected field is proportional to incident field.
The fuze antenna receives this reflected field E,
and converts it to a voltage V r so that

Vr = B'KIej2v [ft ~ + «, (10)

showing that Vr is proportional to /, the trans-
mitting antenna current. The term B' replaces
B and now involves an additional factor trans-
lating field to voltage.

At this point it is necessary to call attention
to the fact that the radio fuzes herein described
and to which the theory we are discussing is
applicable use a common antenna for transmis-
sion and reception and use the same terminals
for transmission and reception. Thus the cur-
rent I in equation (10) represents the trans-
mitter current into the antenna terminals, and
Vr represents the voltage across those same
terminals arising as a result of the presence
of the reflector in space. Since ( Vr/I ) is dimen-
sionally an impedance, we may write

Vr = IZrej2*ft, (11)

where

Zr = B'Kej2* [( -2*/A) +6)' ' (12)

The constants B'K represent the magnitude of
the reflected impedance Zr, and the term ej2^~2x/^

SPECIFICATION OF PROBLEM IN TERMS OF ANTENNA IMPEDANCE

shows the variation of the phase angle as the
distance x to the target changes. In the above
discussion, I has been assumed to be constant
in the presence of the reflector. This assump-
tion is made only for purposes of computing Zr ;
the results obtained hold when I varies, as it
normally does.

Suppose now that the target moves toward
the fuze with a velocity V — dx/dt. Then the
rate of change of total phase 4> of the imped-
ance is

The frequency F with which the impedance Zr
completes its phase cycle is given by

1 d<I> _ 2v
2 nit ~ +\’

(14)

a value identical with the doppler frequency
derived above, equation (7). We thus see that
the reflector can be replaced in the fuze antenna
circuit by a reflected impedance Zr, whose am-
plitude represents the strength of the reflected
voltage and whose rate of change of phase cor-
responds to the doppler frequency shift. In this
derivation of the frequency F, we have neg-
lected relativistic effects; these are, of course,
negligible, just as in the preceding derivation.

For fuzes having a common antenna for
transmission and reception, using common ter-
minals for both, we can represent the behavior

Figure 1. Vector representation of antenna
impedance in presence of reflector.

associated with a moving reflector in the radi-
ation field by the vector diagram shown in
Figure 1.

In this figure Zn represents the impedance

at the antenna terminals in the absence of all
reflectors (free space). Its resistive and reac-
tive components are Ru and X1± respectively.
The term Zr represents the reflected impedance
and Zi the total antenna impedance with the
reflector present. When a target moves toward
the fuze with a velocity v, the end of Zr traces
out a spirallike figure with an angular velocity

= 2tF =

47i -v.

The radius increases as Zr increases.

2'2'3 Approximations Involved in
Impedance Representation

Suppose we consider two systems, each en-
closed in a box with only two terminals avail-
able to the experimenter and no indications
outside to show the contents of the box. Let
box 1 contain a fuze antenna, space for radia-
tion, and a moving reflecting target. Let box 2,
identical in every external detail with box 1,
contain within it a fixed impedance ZX1 and a
variable impedance Zr with magnitudes selected
according to the definitions above.

In a steady-state condition, i.e., with go = 0
and with the fuze in operation long enough for
all transients to die out, no set of measure-
ments can distinguish a difference between the
contents of the two boxes, and they are for all
purposes identical.

If we test the two arrangements by suddenly
applying the r-f voltage to the terminals, there
will be a difference in the way in which the
steady state is reached. This difference is
analogous to the difference in the transient be-
havior of two circuits A and B, where B is
identical with A except for a length of perfect
transmission line attached to its input termi-
nals. If a signal were suddenly applied to the
input terminals of A, a certain transient re-
sponse would be obtained at the output of A.
If the same signal were suddenly applied to the
input of the transmission line attached to B,
the transient response at the output of B would
differ from that at the output of A because of
the delays due to the transmission line. The
steady-state behavior of the two circuits, how-

SECRET

THE RADIATION INTERACTION SYSTEM

ever, would be identical. Thus in the case of
the fuze circuit we can apply the impedance
concept, which is a steady-state concept, to
those cases in which the delays associated with
the radiation link are negligible. We now pro-
ceed to show that these delays are unimportant
in cases of interest.

One effect of the finite transmission time of
the waves is that at any time t the fuze receives
a reflected signal which is characteristic of the
target not at time t but at time (t —r/c), where
r is the distance from target to fuze at the mo-
ment when the signal which arrives at the fuze
at time t started out from the target. This
means that the fuze does not “know” its dis-
tance from the target at any instant, but only
what the distance was at a time (r/c) in the
past. In the region of interest r/c is of the
order of 10-6 sec, during which time the fuze
moves a distance of the order of 10-3 ft. Thus
this effect is seen to be of no importance in de-
termining the position of function of the fuze.
It may be pointed out, however, that for prox-
imity fuzes which work on other principles,
for example, the reflection of radiated sound,
this effect may be of considerable importance.

Another effect of the delay associated with
the radiation link is that it introduces an effec-
tive “time constant” in the fuze circuit because
of the effect which the reflected voltage has on
the antenna voltage, which in turn influences
the reflected voltage, etc. A rough estimate of
the order of magnitude of this time may be ob-
tained by assuming the fuze and antenna sta-
tionary and computing the time required for
the fuze voltage to reach a steady-state value
after being switched on. The time required to
reach equilibrium is assumed for the purposes
of this discussion to be associated entirely with
the propagation of the waves in space and not
at all with delay characteristics of the fuze cir-
cuit itself.

The presence of the reflector induces a volt-
age in the fuze proportional to the voltage in-
duced in the reflector by the fuze antenna. The
above statement can be made more precise by
including the time element; that is, suppose at
time t = 0, the fuze begins to radiate. Some of
the radiation “bounces” back from the reflector,
reaches the fuze again at time A t, usually ap-

preciably less than 10-6 sec, and induces a volt-
age in the fuze antenna. This causes a change
in the radiation; this changed radiation is re-
flected by the target again, and its effect is felt
back at the fuze at time 2 At. This process goes
on until equilibrium is reached.

For the sake of simplicity, assume that the
distance of separation is such that the im-
pressed and reflected voltage in the fuze an-
tenna are always in phase. In this case the
effect of the reflected radiation is to increase
the voltage in the fuze antenna. Let k be the
constant relating the voltage induced in the
fuze antenna by reflected radiation to the volt-
age in the fuze antenna, which was associated
with the original radiation. For many cases of
interest k is of the order of 0.01. Then the vari-
ation in the fuze voltage starting from t — 0
may be represented as in Figure 2, in which
no attempt has been made to represent the true
scale. In the figure F0 represents the voltage at
t — 0. The expression for this variation is

F (t) = F0 + kV(t- At), (15)

which applies for t ^ At. For t < At, V ( t ) = F0.
The equilibrium voltage Vm is the limit of the
series

Foo = F0 (1 + k -f~ k2 -f- /c3 -}- • • •), (16)

- ,~E- <17>

For k << 1, we have

Fro « (1 + k) F0. (17a)

Furthermore

F (A0 = (1 + k) F0. (18)

Thus we see that for small k the first reflection
is responsible for most of the voltage change.
This would be true for any other assumed
phase relation between the impressed and re-
flected voltages in the fuze.

If desired, we may replace the stepwise vari-
ation by a smooth curve, as shown roughly in
Figure 2 ; this smooth curve may be represented
analytically. To do this we replace V (t — At) in
equation (15) by the quantity [V (t) — At
( dV/dt )], the first two terms of the Taylor
expansion.

SPECIFICATION OF PROBLEM IN TERMS OF ANTENNA IMPEDANCE

This gives us the differential equation

m = F„ + *[r(0 - At ^2], (19)

whose solution is

V(t) = - ^ |\ - A-2 *e(1 “ *> - l)HkM) ]. (20)

This equation is, of course, to be applied only
for t — A t.

From equation (20) we find that

and

V(At) = (1 + k) Vo,

agreeing with the previously obtained results.
The order of magnitude of the effect described
above is seen to be quite negligible for the fuzes

K3v0

Figure. 2. Variation of voltage in fuze antenna;
fuze and target stationary.

described here. This effect may, however, take
on fundamental importance for fuzes operating
on other principles, such as those working on
acoustic or pulse-time principles.

2'2'4 Implications of Impedance Concept

The advantage of regarding the basic effect
of a reflector as an impedance change can be
seen when an attempt is made to describe the
phenomenon in terms of another concept which

appears at first plausible, namely, the concept
of the effect of the reflection as a generator e
in series with the radiation impedance Zllf as
in Figure 3. The reflector does indeed create a
voltage e in the antenna. This voltage e how-
ever, changes the current I in the antenna,
ivhich in turn changes e, and so on. This effect
of the change in I upon e must be taken into
account, and the impedance concept does this,
whereas the generator concept as ordinarily
applied does not do so; we do not ordinarily
think of a generated voltage e as being affected
by the current changes which it produces in the
external circuit. Of course, in those cases in
which the reflected voltage e is small enough so
that its effect on / is negligible it may be treated
as a generator.

Another important aspect of the impedance
concept is its essentially geometric character.
It will be shown by more detailed analysis in
the following sections that the reflected imped-
ance in an antenna due to the presence of a
reflector is a function only of the geometric
configuration, of the directive properties of the
antenna, and of the character of the reflector.
The power level at which the antenna radiates
has no effect on the reflected impedance ;
this, of course, is not true of the reflected
voltage. This lack of dependence of the re-
flected impedance upon power level implies that

Figure 3. Generator e in series with fuze
antenna impedance, Z\\.

fuzes with widely differing power outputs can
be made which have the same sensitivity to
reflection. This is indeed true ; fuzes have been
made with radiated power outputs ranging
from % mw to 1 w, with equal sensitivity to
reflection. From the point of view of freedom

SECRE1

THE RADIATION INTERACTION SYSTEM

from interference, however, it is fairly obvious
that the higher power level is desirable. That
is, the reflected voltage increases with the power
level and therefore any extraneous radiation
would have to be so much the stronger to in-
duce, in the fuze, signals comparable in magni-
tude to those coming from the reflector.

2 3 MUTUAL INTERACTION OF SYSTEMS
OF TWO-TERMINAL NETWORKS
INVOLVING RADIATION

In the preceding section it has been shown
that, with certain approximations, the effect of
a reflecting target is equivalent to a change in
the input impedance of the fuze antenna. To the
extent that this is so, the interaction phenom-
ena between fuze antenna and target are de-
scribable in terms of the steady-state analysis
of coupled networks. The familiar concepts of
mutual impedance and reflected or coupled im-
pedance will be used. In fact the antenna imped-
ance change to be evaluated is identical with
the reflected or coupled impedance of circuit
theory.

2,3,1 Fuze Problem as Interaction of
Two-Terminal Networks

For fuzes of the single-antenna type, devel-
oped by Division 4, the problem of the inter-
action with the target reduces to that of com-
puting the reflected impedance. The actual tar-
gets encountered by the fuze radiation may be
relatively simple, as in the case of ground ap-
proach where the fuze can be considered as in-
teracting with its image, or complicated, as in
the case of an aircraft target with its compli-
cated mode of excitation and complicated re-
flection pattern. In the latter case it is custom-
ary toftdetermine performance of a fuze on the
basis of its interaction with a simple target,
such as a half-wave reflecting dipole, and by
experiment to relate the reflection from the
complicated target to that from a simple target.

Thus in the argument which follows in this
section the problem will be set up on the basis
of mutual interaction between a system of n

simple two-terminal networks connected by ra-
diation. In some cases one of these networks
will represent the target antenna. When the
theory has been worked out formally on this
basis, the problem of a complicated target will
be discussed in more detail.

2 3 2 Fundamental Equations

We now formulate the problem in a more
precise way. Let the fuze and reflecting objects
be considered as a system of antennas. If the
ground is involved, we consider it as perfectly
conducting and replace it by the image of each
of the real antennas. For the fuze problem the
fuze antenna and its image are driven ; all other
antennas are parasitic. If some of the other an-
tennas are driven by appropriate generators,
we are then concerned with fuze operation in
the presence of interference or intentional coun-
termeasures. This case is subject to separate
treatment, which is not within the scope of this
volume.

In general, if we have the fuze antenna inter-
acting with 7i — 1 additional antennas, we may
set up n equations :

V\ — IiZn -f I2Z12 + I3Z13 + ' ' ' + InZin,

V2 = I1Z2I + I2Z22 + I3Z23 + * ' * + InZ2n, (21)

Vn ~ IlZnl + hZn2 + 1 3Z n3 + ‘ ‘ * + InZnn,

where 7y is the current in the jth antenna and Vs
is the voltage impressed on the jth antenna.
The set of equations (21) is a well-known way
of representing the interaction between n cou-
pled circuits or n antennas. On account of the
reciprocal relations between antennas, Zti = Zj{.

The meaning of the Z’ s can be elucidated
quite simply. If, for example, we open-circuit
all antennas except No. 1 so that all Us except
/1 are zero, we have Vi = I,Z1U so that Z1X is
the free-space impedance of antenna No. 1 and
Vi and 1 1 are the free-space voltage and cur-
rent, respectively. IiZ2i is the open-circuit
voltage of antenna No. 2 due to current 7i in
antenna No. 1. The term Z2 1 is the mutual
impedance between No. 1 and No. 2. The
input impedance of No. 1 in the presence
of an arbitrary number of other antennas is

MUTUAL INTERACTION OF NETWORKS INVOLVING RADIATION

(Vi/Ii) = Zx. When the n antennas are too far
apart to influence each other, the ZJ s vanish
(i ¥= j) leaving only the Zu’ s.

As has already been mentioned, the ground
is considered as a perfectly conducting plane,
infinite in extent. Modifications required for an
imperfectly conducting ground are considered
later. It is well known that we may “remove”
the ground plane and replace it by images of
each of the antennas above ground. The rela-
tion of the currents in an image and a real
antenna are shown in Figure 4 for two con-
figurations. The arrows point in the direction
of instantaneous current.

If the components of current normal and
parallel to the surface are always as shown, the
boundary conditions at the reflecting surface
will be satisfied and the field of the image above
the plane is identical with the reflected field.

Since each of the images contributes to the
total effect on the fuze antenna, they may ap-
preciably affect the operation of the fuze. When
the target and fuze are far removed from
ground, the effect of their images becomes neg-
ligible. This is essentially true of the applica-
tion of the fuze against enemy aircraft in flight
for fuzes as now constructed. The influence of
the ground in this case will be discussed in
more detail later.

To take account of the effect of the ground
we include the images in the set of n equations,
letting the odd numbers represent real anten-
nas and even numbers the image antennas.
Thus antenna No. 1 represents the fuze, No. 2
its image, No. 3 a real antenna, and No. 4 its
image, and so on, each even number represent-
ing the image of the odd number preceding.

In the notation of equation (21) the bound-
ary conditions will be satisfied if we put

Ir = -/(r- i) (r even).

Since

ZT = Z(r - i), Vr = —V(r- !).

It will now be found that the odd-numbered
equations from equation (21) form a complete
set of n/2 equations to specify the solution for
the n/2 currents in the real antennas. The re-
maining equations can be shown to form an
identical set and so contribute nothing further.

Specific Fuze Equations

In the typical fuze situation only the fuze
antenna is driven. In equations (21) this is rep-
resented by putting V1 = V and V2 = — V with
all other Vi = 0. Let us consider this case and
solve for V/I lf the apparent input impedance Zx
of the fuze antenna. As previously stated, we
use only the odd-numbered equations.

A sufficiently general case which includes all

REAL T REAL

ANTENNA ANTENNA

'7/77777 ^7777777

< ■ IMAGE IMAGE

Figure 4. Relation of currents in real and
image antennas, for horizontal and vertical
cases.

fuze problems of immediate interest arises from
the consideration of two real antennas and their
two images. For this case antenna No. 1 is the
fuze, antenna No. 2 its image, antenna No. 3 is
the target, and antenna No. 4 is its image.

For this special case the appropriate equa-
tions are

V = IiZn — I1Z12 + I3Z13 — IsZu
0 = 1\Z\3 — I1Z2Z + I3Z33 — I 3Z 34

where we have utilized the fact that Zif — ZjV
By symmetry, it is clear that Z23 — Z14. Incor-
porating this in equations (22) we get for the
input impedance Zx of the fuze

v V _ „ v (Z13 - Zu)2

"1 ^11 _ 12 77 7, •

Il 33 — Zj 34

(23)

Equation (23) shows that the impedance of
the fuze antenna is its free-space value ZX1 plus
additional terms representing the presence of
target and ground. Three cases of interest arise.

Case I. Ground Approach. In this case the
fuze uses the ground as a target and antenna
No. 3 with its image No. 4 are absent. This
means that there are no nearby reflectors ex-
cept the ground. For this case Z13 = Zu = 0
and Z 1 reduces to

Z\ — Zu — Z12.

(24)

SECR

THE RADIATION INTERACTION SYSTEM

The coupled impedance is the mutual impedance
Z12 between the fuze and its image. This leads
to an important concept in understanding fuze
operation against the ground ; i.e., in the
ground-approach case the fuze can be thought
of as being fired by its image. Since object and
image are connected by a line normal to the
plane, the vertical distance from fuze to plane
is a determining factor.

Case II. Isolated Airborne Target. It is now
assumed that antennas No. 1 and No. 3 are far
removed from ground in comparison with their
separation. This makes

Z12 = Z\\ = Z34 = 0.

The result is

Z, = Zn ~ f-, (25)

Z/33

and the coupled impedance has the value
(Zi32/Zw). An interesting point should be men-
tioned here in connection with jamming fuzes.
If antenna No. 3 represents a jammer antenna
instead of a target and if Z33 includes some
negative resistance incorporated by feedback
of some sort, Z33 can be made much smaller
than the Z33 obtained if the feedback is re-
moved. Thus a negative resistance jammer will
build up a signal of magnified form and may
cause the fuze to function before it should
normally.

Such a scheme has difficulties of realization
in practice which may make it impossible.

Case III. Airborne Target ivith Ground In-
terference. In this case the full equation (23)
is applicable and must be considered in some
detail. If the target is not moving with respect
to its image, as in the case of a test target, Z34
will be a constant and reasonably small com-
pared with Z33. To a good approximation we
may use Z33 alone. Thus equation (23) includes:

1. Z12 representing the interaction of the
fuze antenna with the ground.

2. Z]32/Z33 representing the interaction of
the fuze antenna with the target plus two other
terms of the same order as this which may
lead to interference.

This is as far as the argument can proceed
without detailed knowledge of the mutual and
self-impedances involved. We now turn atten-
tion to the values of impedance to be expected.

2 4 ANALYTIC EXPRESSIONS FOR
MUTUAL IMPEDANCE, RADIATION
FIELDS ONLY

2 41 Basis of the Argument

We have developed above general expressions
for the apparent input impedance of the fuze
antenna when in the neighborhood of other
antennas, among which may be included the
image of the fuze antenna. These equations will
now be made more specific, so that they can be
applied to actual cases.1’ 4* 9

In the argument to follow we will confine our-
selves to the case of radiation fields alone, leav-
ing the problem of correction due to induction
and quasi-static components to Section 2.10. The
corrections are not necessary to predict fuze
operation in a large majority of cases.

By neglecting the corrections it is possible to
set up a general argument which makes no as-
sumptions about the nature of the current dis-
tribution on the fuze antenna or the mode of
interaction with the reflected radiation. All we
need to know about the fuze antenna is that:
(1) it has two terminals for connection to the
oscillator circuit; (2) when current flows
through these terminals, radiation appears in
the surroundings with a distribution which can
be measured experimentally; and (3) the loss of
energy by radiation appears as a resistance in
the antenna circuit to which the oscillator is
connected.

To derive the necessary expressions we will
first express the field strength E of an antenna
at point P in space in terms of (1) the distance r
from the antenna, (2) the experimentally meas-
ured radiation pattern /(0,</>), (3) the gain G
of the antenna as calculated from (4)

the series radiation resistance Rs, and (5) the
driving point current I into the antenna ter-
minals.

The meaning of Rs may be clarified by repre-
senting the system as in Figure 5, where the
box is the fuze system which emits radiation.
If we integrate the energy flow at infinity when
a current I flows into the terminals, we find that
a certain amount of power is carried away by
radiation. If this power is W, then by definition
2TU 2TT

tls |JJ2 0r JJ*'

FORMULAS FOR MUTUAL IMPEDANCE, RADIATION FIELDS ONLY

Now there may be other components in the box
which dissipate energy. They are not included
in Rs.

If we measure the input impedance at the
terminals TT when the box is in free space in

Figure 5. Representation of fuze system emit-
ting radiation.

the absence of reflectors, the result is Zu. When
reflectors are present, the result is Z u as defined
previously. The above definition of Rs implies
that all the antenna current I flows through Rs ,
meaning that Rs is in series with I. Likewise
the coupled impedance representing the reflec-
tor will also be in series with I. The argument
upon which equations (21) are based then
means that we consider the antenna as equiv-
alent to the circuit in Figure 6. Thus ZX1 — Rs +
Ra + jX8 and A Z represents the reflected im-
pedance. The term Ra represents the ohmic
losses in the antenna, which are quite small and
will be neglected unless otherwise specified.

When the field relations have been derived, it
will then be necessary to determine the response
of the antenna to radiation falling upon it. This
will be derived with the aid of the reciprocity
theorem. The two concepts will serve to solve
the fuze problem in so far as pure radiation
fields are concerned.

Field Equations for Arbitrary
Antenna

We assume a spherical coordinate system,
with the origin at the center of the antenna
and the antenna lying along the polar axis. The
electric field strength E is a vector function of
position. If we describe a large imaginary
sphere around the antenna, then a plot of the
field strength E, on the surface of the sphere,
as a function of the polar angle and the azimuth
angle <f> is known as the space radiation pat-
tern of the antenna. If we normalize the values

of \E\ found around the sphere so that the
maximum value is unity, the dependence on
6 and <£ is known as The actual value

of the field strength at any point ( 0,<f> ) on the
sphere is given by

\E\ = £0/(W), (26)

where E 0 is the maximum value of \E\ on the
surface of the sphere. We assume that /(0,<£)
has been determined experimentally (see Sec-
tion 2.8) .

Figure 6. Series equivalent circuit of fuze
antenna.

The power W radiated through the sphere is
obtained by integrating the Poynting vector
over the surface of the sphere and is given in
mks units by

2jt n

w = r¥o f f p (®**> sin md4, (27)

where

Ed2 f2
2Z0

(28)

2tt 7 r

f f r si

sin dddd(t),

and

Z o = V (mAK),

the “intrinsic impedance” of free space, \i and K
being the permeability and dielectric constant
respectively of free space, or air. The term
Z0 = 120t r ohms.

Taking into account equations (26) and (28),
we write

E = \ {/(0,4>yM ■ <2”/X)1}, (29)

where l is the wavelength. This expression ig-

SECRET

THE RADIATION INTERACTION SYSTEM

nores a possible additive phase shift which may
be a function of It will be introduced

when needed.

We may now introduce the concept of gain
of an antenna. If we compare two antennas,
each of which radiates so as to produce equal
values of E0 at a given distance r, then the an-
tenna which radiates less power has the greater
gain G. An antenna for which the space radi-
ation is spherical, i.e., one which radiates
equally in all directions, has the lowest possible
gain. From equation (28) we see that for two
antennas, No. 1 and No. 2, with equal values
of E0 at the same value of r

G* _ Wi _ 7i

Gi W 2 72

(30)

For an isotropic radiator y = 4tt. If we arbi-
trarily assign this antenna a gain of unity, we
have for any antenna

4 7T
7 '

(31)

Typical values of G for representative antennas
will be found in Figures 21 through 24.
Equation (29) may now be transformed:

E = \ - (2"A)1}- (32)

As already indicated, we put

Ra = JJJ2 , (33)

and rewrite equation (32) as

E = 7 ypff- {m^VC -2'rA)}( (34)

where G = |/| e,ut. The factor j correctly re-
lates the phase of E to that of h in the case of
an elementary dipole. For other antennas, there
may still be an additional phase shift, as men-
tioned above. This is the final equation relating
the field to the antenna and shows the radiation
field as a function of position around the fuze
antenna. To solve the fuze problem we need to
know how this arbitrary antenna responds to
fields as a receiver. A discussion of the problem
follows.

2 43 Mutual Impedance Between Two
Arbitrary Antennas

In the following discussion we assume that
the radiation is in the form of plane waves.
This in effect means that the absolute value of
the field does not vary over the length of the
antenna for distances at which we are inter-
ested.

We know that a current in one element sets
up a voltage in another. These may be coupled
by radiation, in which case the radiation field
from one antenna carrying current h generates
a voltage in the other and we mgy say that the
impressed field on an antenna generates a volt-
age at its terminals. Since the antenna is a
linear circuit element, we can say that the volt-
age at its terminals is proportional to the field
intensity acting upon it. If this field varies
along the antenna, it will be necessary to pick
some reference point in space and say that

V = IE, (35)

where E is the value of the field at this refer-
ence point and l is a constant of proportion-
ality having dimensions of length, usually called
the effective length. The term V is the open-
circuit voltage at the antenna terminals. It is
customary to select the feed point of the an-
tenna as the reference point. If this is done,
E will then represent the field intensity at this
point in space if the antenna is assumed to be
absent while the field intensity is determined.

Now consider two arbitrary antennas, No. 1
and No. 3, like those treated in Section 2.4.2,
separated by distance r with currents h and /3
at the feed points. Assume that h gives rise to
a voltage V3 at the terminals of No. 3 when /3
is zero, and assume that /3 gives rise to a volt-
age V1 at the terminals of No. 1 when h = 0.
By means of equations (35) and (34) we can
write

J' 3 = h -1 yj~] ^ Vi^, <£13) (cos r)jeA 2’rr/X),

(36)

Vi = h £ (cos r)je“

Here we introduce the angle x to take account
of any skew relation between the two antennas.

ANALYTICAL FORM OF REFLECTED IMPEDANCE

Here fi(613y<f>i3) denotes the value of f(Oy<f>)
for antenna No. 1 in the direction joining the
fuze and target antennas No. 1 and No. 3 re-
spectively, and f3(031y<j>31) has an analogous
meaning.

At this point we shall call upon the Rayleigh-
Carson reciprocity theorem. The statement of
this theorem as given by Carson is as follows :95

“Let an emf E /, inserted in any branch, des-
ignated as No. 1, of a transducer, produce a
current /2' in any other branch, No. 2; corre-
spondingly, let an emf E2" inserted in branch
No. 2 produce a current //' in branch No. 1;
then I/'E/ — I2'E2".”

A transducer is defined as “a complete trans-
mission system which may or may not include
a radio link, which has accessible branches,
either of which may act as the transmitting
branch while the other acts as the receiving
branch.”

The theorem of reciprocity applied here
means that

F3 Vi „

T = T = Zl3

O 1 3

= lj (COS r )je* ~ 2"A).

(37)

Equations (36) and (37) give

h /i (^13,013) = h °4^3 3 f 3(031, <f>n)j

(38)

h _ h

IZqRssGs , f \ Iz0rsiGl { , \

V — 4^ — j 3(031,931) yl — ^ JlWl3,<Pl3;

Thus for any antenna we may write

ZqR8G

M+)

= c

(39)

(40)

where C is a constant not involving any of the
variables in equation (39). Finally

l = C

(41)

showing that as a receiver the antenna behaves

the same as a transmitter in its dependence on
Z0 Rs, G, and

The mutual impedance between two arbitrary
antennas can now be expressed. Antenna No. 1
impresses a field on antenna No. 3 as given by
equation (34) ; antenna No. 3 receives it with
an effective length l3 given by equation (41).

7 V3 GEl CZ0 /~d e> p ri

^13 ~ ~T ~ ~f — — ~A V .ttsi/£S3CriCr3 •

1 1 1 1 4 TTf

/i(013,<M /3(031,03l) (cos r)jej( - 2vr/x\ (42)

If we can evaluate the constant C for any two
antennas, we have it for all antennas. In Sec-
tion 2.14 it is shown that C has the value
(2k/ Z0) . Inserting this into equation (42) gives

Z13 — 2^ a/ RslRssGiG3 fi(du,4>u)

f 3(631, <t>3i) (cos r)jej( “2irr/x). (43)

Equation (43) represents the mutual imped-
ance between two arbitrary antennas separated
far enough so that the radiation field (1/r term)
is the only one of importance.

We have seen in Section 2.3 that the antenna
impedance of the fuze in the presence of reflect-
ing targets can be represented as the sum of its
self-impedance in free space plus terms in-
volving mutual impedances Z{j and the self-
impedance of the reflectors. Equation (43)
gives the analytic form of the mutual imped-
ances Zijy if 1 and 3 be replaced by i and j,
respectively.

We are now in position to apply this general
formula to special cases representing a fuze
approaching ground (interacting with its
image) or a fuze approaching an airborne tar-
get well away from the ground.

2 3 ANALYTICAL FORM OF REFLECTED
IMPEDANCE9

The analytic expression equation (43), de-
rived in the preceding section, will be applied
to three special cases and appropriate working
formulas discussed. The general properties of
the reflected impedance common to all three
cases will then be discussed.

b Bibliographical references pertinent to this section
are 1, 13, 16, 17, 22, 27, 51, 53, 93.

secretI

THE RADIATION INTERACTION SYSTEM

The general equations for the total antenna
impedance of the fuze discussed in Section 2.8.1
were applied to three special cases with the
following results :

Case I. Ground Approach.

Z ! = Zn - Z12. (24)

Case II. Airborne Target Far From Ground.

Zx = Zn - (25)

Case 111. Ground Interference Case.

Zx = Zxl - Zx2 - (f13 ~ yu)'. (23)

6 33 — ^34

Each of these equations is of the form

Z\ = Zn — Zr, (44)

where Zv represents the reflected impedance.
The vector interpretation of this equation has
already been given in Figure 1.

Ground-Approach Equation

A fuze approaching ground, in the absence of
other reflectors, is interacting with its image
and Zr — Zi2. Furthermore, since antenna No. 2
is the image of antenna No. 1, we see that

Rsl = Rs2)

Gi = G2,

fl (012,012) = f2 (021, 02l),

T = 0,

r = 2h (h = height above ground).

Introducing these relations in equation (43),
we have

Zr = Zxx = ^ GR.fi 2 (0,2 ,0,s)je(-*'*A). (45)

Equation (45) gives the detailed form of Zr
for approach to a perfectly reflecting ground
of large extent. As in equation (29) and sub-
sequent equations a possible additive phase
shift is ignored. From the results obtained in
the case of the perfect reflector, we may ex-
trapolate to actual grounds (plane reflectors)
by the use of a reflection coefficient n. For pur-
poses of these applications, the effective reflec-
tion coefficient n for a given surface is defined
in such a way that the signal magnitude re-

ceived by a fuze circuit, because of the presence
of the surface, is n times the signal that would
be received from a perfect reflector in the same
position as the actual reflector.

This definition was set up to avoid possible
errors in using the reflection coefficients de-
rived for plane waves on the classical theory.
(The equiphase surfaces of radiation from the
fuze antennas have appreciable curvature at the
usual distance of interest in fuze applications.)
As a matter of fact, however, it has been found
that the values of n found according to the
above definition agree well with the published
values of n based on the plane wave theory and,
to the accuracy needed for fuze calculations, are
independent of the height above the ground.
Additional comments regarding n are to be
found in Sections 2.9 and 2.14.

The reflection coefficient n has been measured
by moving a fuze over a perfect reflector then
over ground and comparing results (see Sec-
tion 2.9) .

When the reflection coefficient is included,
we have

Zr = l ^ GRs fl2 (0i2,0i2)^(->4^/x). (45a)

Airborne Target Equation
For target and fuze a long way from ground

With the aid of equation (43) we get
* ■ -&)'■

^Slfis3b?lG?3/l2(013,013)/32(031,03l)CQS2T ^ -j^r/ \ (4Q)

Z33

Equation (46) is limited in its application to
cases where the target can be considered as a
single antenna with a single feed point. Thus it
represents the case for a dipole reflector or a
strip of “window.” If the target can be repre-
sented as an array of simple antennas, then Z,
would involve mutual interaction with the whole
family, including terms arising from the mutual
interactions between members of the array.

If the target is not made up of linear anten-

SECRET

ANALYTICAL FORM OF REFLECTED IMPEDANCE

nas but is a geometric shape capable of excita-
tion in a complex manner, equation (46) cannot
be used as it stands, since it is obvious that we
cannot cut a complicated target arbitrarily and
reproduce its complicated current distribution
by feeding it at one point.

If we knew the current distribution on the
target arising in response to the radiation from
the fuze, we might proceed as follows: (1) find
the number and location of the feed points nec-
essary to reproduce this distribution, (2) deter-
mine for the target when excited by

each feed alone, (3) treat each feed with its
f as a single antenna, (4) measure the
mutual impedance between feed points, and
(5) proceed with the general equations (21).
For any ordinary target such a process is im-
possibly complicated, and we resort to more
tractable methods.

Again we assume the fuze and the target to
be far enough apart so that we can consider the
fuze radiation to consist of plane waves at the
target. We then determine the reflecting power
A of the target as follows: (1) We irradiate
the target with plane waves with a field inten-
sity Eif and (2) we measure the field Er reflected
back along the direction of the incident radi-
ation. If this field Er is measured at distance r
from the target, A is defined by

Er = ^4 (47)

The (1/r) dependence of Er is consistent with
our initial assumptions concerning the plane
waves from the fuze. Note that A has the di-
mensions of a length.

For a single linear antenna, it follows from
the definition of A that A for such an antenna
is given by

A = 2^ G3t/*32 (03i,$3i) cos t. (48)

As an example, consider A for a resonant half-
wave reflecting dipole oriented for maximum
reflection. In this case Rs 3 = Z33; G3 = 1.64,
/32 = 1 cos x = 1, and A3 = 0.26. Typical
values of A for other simple reflectors are given
in a paper by Mott.93 In particular,

A (sphere) = Ja, where a is the radius of the sphere;

Z/2

A (flat sheet) = — where L 2 is the area of the sheet.

We now express Zr in terms of A as defined
above with the aid of equation (48) and get,

Zr = A ~ fl.A/i2 (0„,<fos) (cos (50)

For each antenna, x is the angle between the
plane of polarization of the incident radiation
and the plane formed by r and the axis of the
antenna.

If the antenna is a complicated structure,
the meaning of A in equation (50) will require
modification to include effects of the twisting
of polarization of the incident radiation.

In the case of an actual aircraft target it
would be necessary to know A at all angles,
since the fuze sees a continually varying aspect
as it approaches the target. Thus the calcula-
tion of Zr by the use of equation (50) would
require analytic expression of A as a function
of direction toward the fuze.

The necessary information can be achieved
in a more expeditious manner. An actual fuze
is set up and the target moved past it slowly
while the signal in the fuze is recorded. The
recorded wave can be reproduced and used
directly for testing fuze circuits. Such experi-
ments are described in detail in Section 2.11.

To relate these measurements with our cal-
culations the strength of the reflection was com-
pared with the reflection from a resonant half-
wave dipole, which can be computed directly
from equation (46), giving

Zr = 0.042^y RalGJ{2 (0,3,<*«i3)e-jW\ (51)

for the dipole orientation which gives maxi-
mum reflection.

In general we find that the maximum reflec-
tion from the aircraft as it passes the fuze is
N times the reflection from a dipole given by
equation (51). It has the same dependence
upon distance as a dipole for approaches that
are not too close. The term N will not be a con-
stant for a given target but will depend on A.

From Mott’s paper93 we find that A for a flat
sheet of area L2 is L2/A, and A for a dipole is
0.26A. Thus, a sheet of area L2 is the equivalent
of N dipoles, where

„ 3.88L2

A2 '

(49)

THE RADIATION INTERACTION SYSTEM

With the aid of equation (51) this shows that
Zr is independent of \ for the case of a flat sheet.

253 Ground Interference

The general equation covering this case for a
fuze and one target is given by equation (23) :

Zi = Zn - Zn - (y13 ~ y“)2. (23)

The detailed treatment of this case for a
complicated target is beyond the scope of this
report. However, certain general properties can
be observed.

The symbol Zr consists of two terms, the first
representing the reflection from the ground and
the second the reflection from the target, in-
cluding the effect of the images. Now, in gen-
eral, Z34 is small compared with Z33 ; it will be
10 per cent or less for a dipole if the separation
is 4 l or more. Thus we may write Zr as

7 7 , (Zu ~ ZU)2

Zjr ~ Zri2 ~\ ,

^33

with reasonable accuracy in so far as absolute
magnitudes are concerned.

When the distance between antennas No. 1
and No. 3 is large compared to the distance
between No. 1 and No. 2, |Zi3| is nearly equal
to \Z14\ and the effect of the target is compli-
cated by phase relations between Z13 and Zi4,
giving rise to interference in the reflection
which may be quite pronounced. However, when
the fuze gets close to the target, or when the
distance from fuze to target is much less than
the distance between target and ground, Zu and
Z12 are small, and the signal is approximately
equal to the free-space signal.

The situation is further complicated by the
directional properties of target and fuze. In
each impedance Zijt f(6,<j>) must be evaluated
in the direction (0^,0^) from each antenna.

Any further discussion must be limited to
special cases. One particular example is of in-
terest. Neglecting directional factors, we com-
pare the strength of the reflection from ground
with the reflection from a resonant half-wave
dipole oriented for maximum reflection to the
fuze. We wish to determine at what distance r

from a dipole the reflection is the same as from
the ground at distance h.

From equations (51) and (45) \Z12\ = \Z13\
when

0.042^/^ (ft,,*.) = |^/i2 (012, <M.

Now if the radiation pattern of the fuze be such
that

fl2 (013,</>13) = /l2 (012,012),

then the signals are equal when r = V 0.52 \h.
If X is about 10 ft and h is 10,000 ft, then r =
230 ft.

In the early days of fuze design this limita-
tion caused some needless concern. In the first
place the reflection from an airplane is of the
order of 10 times that from a dipole. In the sec-
ond place the radius of action of practical fuzes
described in this volume is about 75 ft. In the
third place the orientation with respect to the
ground and the relative motions involved make
the ground signal less important.

Only in special cases where the fuze is used
against airborne targets near the ground does
the ground reflection become a limitation on
fuze operation.

254 Special Considerations of Transverse
Antenna Fuze

In the preceding discussion it has been as-
sumed that the fuze can be represented by a
single antenna. This applies for fuzes using the
missile as the antenna. In the cases of fuzes
with transverse dipoles as antennas (T-51 and
T-82), the expected variations of antenna im-
pedance are complicated by the presence of the
body of the missile near the fuze antenna. If the
transmitter and receiver circuits are not elec-
trically and mechanically balanced with respect
to the missile, longitudinal currents are excited
in it and these radiate energy. As a result there
is in effect an additional antenna in the system,
and its contribution to the performance of the
fuze must be considered. Furthermore, even if
perfect balance is obtained, the missile serves
as a director or reflector behind the fuze to alter

ANALYTICAL FORM OF REFLECTED IMPEDANCE

its sensitivity pattern (see patterns in Section
2.8). We are not here concerned with this latter
effect. We are concerned with the results of in-
cidental unbalance that arises in the manufac-
ture of fuzes.

To study them we idealize the system as two
thin antennas arranged at right angles and con-
cern ourselves with the reflected impedance Zr
when this system approaches the ground. To
the extent that we can represent the system by
two thin antennas the general arguments of
Section 2.8 can be applied.

The arrangement to be considered is shown in
Figure 7. «

GROUND PLANE

Figure 7. Representation of transverse dipole
and projectile, with their images.

Antenna No. 1 is the fuze antenna. This case
was treated in Section 2.3 for another purpose
and led to equation (23), which is

Zl = Zn - Zn ~ (y3 ~ |h)2. (52)

Z/33 — Z/34

To interpret this equation we expand and get

Zi = Z 11

Z3
+ 2

3 — Z34

ZnZu
Z 33 — Z 3.

Z33 — Z 3,

(53)

The first two terms of equation (53) represent
the interaction of the T-51 or T-82 fuze with
its image in the absence of any vehicle. Zri rep-
resents the free-space impedance and Z12 the
reflected signal. This has been generally inter-
preted as the actual working signal in the T-51
fuze when used. If the balance is perfect,
Z13 — 0 and equation (53) reduces to

which shows that even though the projectile is
not excited directly by the fuze antenna it
nevertheless contributes to Zr. In general ZM is
small compared to Z33 and serves to modulate
the second-order reflection terms. We will neg-
lect it in comparison with Z33 in the remainder
of the discussion. Also we may use Z33 and Z4 4
interchangeably, since they are images of each
other. The term (Zi42/Z44) represents the re-
flection from the image of the projectile as a
target for the fuze antenna.

To discuss the problem further we need a co-
ordinate system. We choose the z direction as
the axis of the missile with x and y axes per-
pendicular to it, the x axis being the axis of the
transverse dipole. We also choose a as a polar
angle. It is the angle between z and the normal
to the ground, that is, the striking angle of the
projectile referred to the vertical. The term 5
is an azimuth angle measuring the angle be-
tween the x axis and the plane including the
axis of the projectile and the normal to the
ground (plane of incidence). To estimate
the order of magnitude of this effect we shall
make the further assumption that antenna
No. 3, representing the projectile, is a resonant
half-wave antenna. We shall also consider the
radiation pattern of such an antenna to be
f(0) — sin 6, when 0 has the meaning pre-
viously assigned ; this is a good enough approx-
imation to the true pattern for this argument.

The field components from the x-axis dipole
will be

Er = 0,

Ea = ki cos a cos 8, (55)

E 8 = ki sin 8,

where

*1 = 7 A. (56)

The component Ea will be in the vertical plane
containing antenna No. 4 and will give rise to
a voltage in it. The term E& will always be per-
pendicular to the plane and will produce no
voltage in antenna No. 4. Thus Z44 will be

„ _ Eg U

•£14 — j ,

Z\ — Zn — Z12 — ~ 14 7 , (54)

Z/33 — Zj 34

U COS a COS

(57)

SECRET

THE RADIATION INTERACTION SYSTEM

. 2A IRsiGi RsaGi .

Z 14 1 = — ^ — -j— cos a COS o sin a. (08)

We then find

4A2 RsiGiGa •> s • 9 /-n\

/ r cos- a cos- 5 sin- a. (o9)
r2 (4tt)2

when we assume the projectile is resonant, so
that Rs 4 = Z44. This will give rise to a change
(Z,.)4 1 in antenna impedance in antenna No. 1
given by

= RsiGiGa cos2 a cos2 5 sin2 «.
47 rr

(60)

We compare this with the so-called normal im-
pedance change

i Zu\ = RsiGi (1 — sin2 a cos2 5). (61)

27rr

The worst case we will be interested in will be
8 = 0, a = 45 degrees, and

|Zl2I = 2 VrR,lGl>’ (62)

l(Zr)i\ = R,iGiGi. (63)

The signals arising from Z12 and (Zr)4 will
have an unknown phase relation depending

upon striking angle and the size of the pro-

jectile. We consider the worst cases where they
may be in or out of phase. The interference will
change the response by the ratio

(A/47rr) G, ± (\2/167rV2) G1G4 1 . 0.13A

(X/4irr) G1 1 ± “• (64)

For heights of operation of r = 10A (that is,
h = 5A) the maximum change in reflected sig-
nal will be approximately ±1.3 per cent. For
other angles a < 45 degrees, 8 0 degrees, the

correction will be less, being 0 for 8 = 90 de-
grees and all values of a.

We thus conclude that the variations in height
of burst from the source are small for a per-
fectly balanced transverse antenna.

A greater source of error is the unpredictable
value of 8. For a = 45 degrees the reflected sig-
nal changes from 1 to % as 8 varies from 90 to
0 degrees.

We now turn our attention to the correction

arising when the antenna is not perfectly bal-
anced. In this case Z13 ^ 0. There will be two
terms of interest.

4 - Zl*2

Z 33 '

(65)

D 2ZuZu

B = 7 •

" 33

(66)

The term A is a fixed term independent of
r or h and shows merely the amount of reflected
impedance in antenna No. 1 by virtue of its ex-
citation of antenna No. 3 by some unbalance.
This term, being constant, will give rise to no
signals. It is merely a measure of the coupling
between antenna No. 1 and antenna No. 3.

The term B gives rise to an additional signal.
As before we will consider only absolute values
and disregard relative phases, since the abso-
lute values will indicate the maximum value of
the corrections that may arise.

To estimate the coupling we note that IiZ14
represents the free-space voltage at antenna
No. 1. If Z13 is small we can say also that hZls
represents the voltage coupled into antenna
No. 3. We define k by the relation

7 \Z\ 3

Z 13

IiZu

zli

Experiments have shown that k is of the
order 0.01 for well-balanced fuzes and may be
as large as 0.1 for very poor balance, so poor in
fact that the arrangement would never be used.
These values are based upon the center point
of the parasitic antenna as the reference point.

Now 'Zn | is about 300 ohms and |Z33j >73
ohms. Hence |Z13| is approximately 3 ohms and

1 B

JX3

Z14

— ft nor pont rvf

Zi4

I Z\2

- 73

Z12

O Lcll t DI

Z 12

For all angles of approach that are of interest
|Zi4j < Z12 1 so that the correction is less than
10 per cent. If the unbalance becomes large this
correction becomes sizable and can lead to a
considerable change in function height. In most
cases the projectile is nonresonant and |Z33| is
considerably greater than 73 ohms. Thus inci-
dental unbalance is not so important when the
projectile is nonresonant. When resonance is
approached the response becomes critical to
unbalance as has been experimentally observed.

SECR

ANALYTICAL FORM OF REFLECTED IMPEDANCE

2,5 5 General Properties of the Reflected
Impedance

We are primarily interested in the two basic
equations, the ground-approach equation and
the airborne-target equation. We repeat them
here for convenience.

47 rh

GJts.Pidn, *12)

(ground approach) , (45)

and

RsiG&i* (013, *18) (cos T)e-**'A

Z7rrz

(airborne target), (50)

or in alternative form for a linear antenna re-
flector

(cos2

Zaa

(46)

It should be remembered that the phase factor
should carry an undetermined constant phase
shift, related to the antenna properties of tar-
get and fuze.

f2 (0,*) . Thus the radiation patterns become
characteristic of a given vehicle, and Rs can be
adjusted to match the transmitter properly
with the assurance that Rs will have little effect
on Zr/Rs.

A particular example is most enlightening.
For antennas whose length lies in the range
0 X/2, G varies from 1.5 to 1.64, about a 10

per cent change. The term /2(0,c/>) in the
worst direction only differs by 15 per cent in
the two extreme cases. Thus the quantity
Zr/Rs is practically independent of antenna
length in this range. On the other hand, Rs
varies from 73 ohms for a half-wave antenna
to zero [as (LA)2] for short antennas, where
L is the length of the short antenna. We can
thus state generally that all fuzes, whose re-
sponse to a fixed value of (Zr/Rs) is the same,
will have practically the same response to a
given target no matter what the length of the
fuze antenna is, provided it is considerably less
than a half-wave long. The statement is also
true for all loop antennas whose dimensions
are small compared with l.

For longer antennas /2(0,*) is sensitive to
l, and each must be considered as a special case.

Doppler Frequency

As seen in Section 2.2 the phase has a fre-
quency F = (2v/l), if | ( dh/dt ) | or j ( dr/dt ) j
be replaced by v. This is identically the doppler
frequency.

Dependence upon Rs

Let us write Rs for Rs i. We note that Zr is
proportional to Rs, as would be expected, since
it is the power dissipated in Rs that accounts
for the radiation fields which make the inter-
action possible.

It will be observed in a later section that the
dimensionless quantity (Zr/Rs) is most con-
venient for assessing fuze circuit response. It
exhibits the effect of a reflector as a fractional
change in antenna impedance which depends
only upon the nature of the target and the
directive properties of the fuze antenna (which
are relatively independent of Rs).

It has been found experimentally that small
changes in the antenna feed point can change
Rg over a wide range with almost no effect on

Dependance on Distance

The magnitude of (Zr/RJ depends upon the
distance through the dimensionless ratio
r/l or h/l. This means that 1 is a scale factor
in determining fuze performance. Response to
the presence of a target is determined by the
number of wavelengths in the distance to the
target; thus, for example, the signal received
from ground reflection at a given height is
greater for greater a.

Dependence on Direction to Target :
Definition of Directivity

The size of Zr/Rs is proportional to Gi/i2
(0i3, *13). Now Gi is a constant for a given
fuze so that /i2(0,*) tells how well a fuze
“sees” targets in various directions. The term
/i2(0,*) is the power radiation pattern of the
fuze antenna ; it is called the directivity pattern
in this report. A plot of /2(0,</>) will show
(other things being equal) the distance at
which a fuze will function upon approach to a
target.

SECRET

THE RADIATION INTERACTION SYSTEM

Now it will be observed that

GJiKtaju) = Wu _ F(ei3,<t>n), (67)

47T W

where JE13 is the power radiated per unit solid
angle in the direction (^13,^13) and W is the
total power radiated. Thus F(0i3,<£i3) repre-
sents the fraction of the total power radiated
per unit solid angle in the direction (#13, <£13) •
Thus f2 (Ois,<fnz) is an indication of how effi-
ciently the total radiated power is used.

Typical directivity patterns are described in
Section 2.8.

Independence of Power Level

It is clear, as was anticipated in Section 2.2,
that Zr/Rs does not depend upon the power level
at which the fuze radiates.

26 CIRCUIT RESPONSE TO ANTENNA
IMPEDANCE MODULATION

Series and Parallel Expressions
for (Zt/Rs)

Differential Signals

The foregoing analysis has been based upon
the concept of series antenna resistance and re-
actance. In actual cases, however, it is often
more convenient to deal with the equivalent
parallel quantities. We therefore proceed to de-
rive expressions for the changes in parallel
antenna resistance and reactance due to a re-
flector.

We deal with the two circuits in Figure 8.
The terms and \X8, or \XV and A Rp, are the
changes in antenna resistance and reactance
resulting from the presence of a reflector; Rp
and Xp, or R8 and XH, are the free-space values.

It is easily shown that in the absence of the
incremental quantities we have

Rs 2 + X82
Rs 9

Rs 2 + X*2
X8 '

(68)

(69)

consider the differentials of Rp and Xp as equiv-
alent to their increments.

Then

dR„ = || (RS - X,2) + Jr (2 S,R,), (70)

and

dXp = dR, + ~s (X,2 - R?). (71)

By defining Q = X8/Rs and by appropriate
manipulation we find

dRp _ dRs (1 - Q2\ dXs (2 Q \

' Rp ~ Rs\ 1 + Q2) ^ Rs \l + Q2)’ V ;
and

dXp 1 r dRs • 2Q dXs(Q* - 1)1 ,

Q [^(1 + Q2) RS(Q2 + 1) J*

Now as we have seen, dRs and dXs are the
components of a vector Zr which may be writ-
ten as

Zr = \Zr\e*, a = (74)

where x represents the distance r or h from
fuze to target. The term 8 depends on the par-

SERIES CIRCUIT PARALLEL CIRCUIT

Figure 8. Series and parallel equivalent fuze
circuits.

ticular case being discussed, but is unimportant
for our present purposes. It will be found con-
venient to use the dimensionless ratio ( Zr/Rs )
which we define as a new vector M and write

M= M 0eja,

where

For small increments of impedance we may

(75)

CIRCUIT RESPONSE TO ANTENNA IMPEDANCE MODULATION

If we define an auxiliary angle (3 by the relations
1 - Q2

sm 0 =

cosjS =

1 + Qv

2 Q

l + Qr

We can rewrite equations (72) and (73) as

TT2 = M 0 sin (a - /»),

xL p

dX p M o / Q\
T7 “ -Q cos (a -

(76)

(77)

For purposes of convenience, it may be de-
sirable at times to work in terms of the admit-
tance Y, which is defined through the relation

Y =

Rs + jX.’
Rs

Rs2 + Xs2

Y = G — jB,

Rs2 + Xs2’

(78)

where the conductance G and susceptance B are
seen to be

G =

B =

Rs2 + X*2’
X,

(79)

Rs2 + XX

From equations (68) and (69) it is seen that

conductor, the impedance Zr becomes a sizable
fraction of Rs. We need expressions for the cor-
rections that may be needed in interpreting
such tests. By replacing Rs by ( Rs + A#s) and
Xs by ( Ms -f- AXS) in equations (68) and (69),
we get

it = r+i„cosa[MoSin(a +/3) + rriXf“2}

(82)

it = . ~ Mo . [gM«c°B(«+0 + rtwM

1 + -Q-sina L

(83)

It will be observed that these equations re-
duce to the differential forms when M0 is small
enough. For larger signals the equations indi-
cate the presence of a “d-c shift” which is small
when Q is large. They also show that equation
(76) is in error by a fraction equal to M0 even
for very large Q. Thus in tests where M0 is 10
per cent the field measurements are accurate to
10 per cent and can be corrected if desired.
Some caution must be used in applying equa-
tions (82) and (83) to an actual case since in-
duction fields will also contribute to Zr at about
the same separation that leads to large M0.

However, in most fuze applications M0 is
about 0.5 per cent, and the differential forms
have ample accuracy.

and

G = k

(80)

Therefore

dRp

and

(81)

dXp

Xpm

Finite Signals

It will be observed later in the chapter that
the actual working signals when the fuze is
operating normally are so small that the differ-
ential representation given above is completely
adequate. However, when tests are made on the
fuze by measuring its response close to a large

Specification of Fuze Circuit
Parameters

It has been shown that both the resistive and
reactive components of the antenna are altered
by the presence of a reflector. To make a work-
ing fuze it will be necessary to devise a circuit
which will respond in some manner to the
change in antenna impedance. Such circuits are
described in detail in Chapter 3.

For purposes of further analysis we assume
that the voltage or current in some part of the
fuze circuit changes in response to the antenna
variations and that this change is used to actu-
ate the fuze. We will call this particular fuze
parameter V, representing a voltage, although
it might as well represent a current. In order
to continue the discussion of the antenna prob-

THE RADIATION INTERACTION SYSTEM

lem, we assume that the behavior of the circuit
is known and that

V = f (In Rs, In X8) = g (In Rp, In Xp), (84)

where for purposes of convenience we express
the functional relationship in terms of the natu-
ral logarithm of the impedance components.
Thus

(85)

We define

o _ ^ c SV

p d In Rp ’ s ~ d In Rs'

T = dV • T - dV

p d In Xp’ s ~ d In AY

(86)

The quantities Sp, Tp or their corresponding
transforms Ss, Ts describe the behavior of the
fuze circuit when the antenna impedance varies.
Ss and Sp are called the series and parallel re-
sistance sensitivities respectively and Ts, Tp are
called reactance sensitivities in a similar man-
ner. All four quantities are functions of X8, Rs
or Xp , Rp which make up the free-space input
impedance Z0 of the antenna. In Chapter 3 the
values of these quantities for typical circuits
will be derived.

With the aid of the definitions of equation
(86) we may write

r» dR s . rp dX s

^ Ts X?

or dR p „ dX p

u ' 1 p y ’

n p p

M o (Ss cos ^ sin a) ,

(87)

o[sp sin (a

Tp _

688)

Mo | Sp sin (a + 0) + ^ cos (a + /S) J.

If we make use of the complex notation and
always consider dV to be the real part of a cor-
responding complex quantity, we may write

dV = MS = MoS0eX° ~ *>,

where

*-(*•- if)

(89)

S = ^.Spsin/3 - “cos0^ - j^S pcos0 - ^sin/3^:

(90)

(91)

tan 7)

T 2QSP - ^ (1 - Q2)
QSs = Sp (1 - Q2) + 2 Tp '

(92)

The appearance of the terms TJQ and Tp/Q
in equations (91) suggests redefinitions of Ts
and Tp to include Q. This can be done (see
Chapter 3), but Ts and Tp so defined will then
not have the logarithmic form commonly used
for Ss and Sp. We keep the Q to maintain sym-
metry with the commonly accepted notation.

Since we are dealing with different mathe-
matical representations of the same antenna
the voltage change dV will be the same no mat-
ter which representation is used. This was im-
plicit in equation (89).

The basic equation (89) represents in simple
form the response of the fuze circuit to a mov-
ing target. While q is a fixed quantity for any
given fuze, a (=r —4? tx/\ -f- 8) varies with fuze-
target separation and therefore with time. The
voltage change is seen to be proportional to M0
and to S0, the r-f sensitivity. If the fuze has re-
actance sensitivity as well as resistance sensi-
tivity, both contribute to S0 and give rise to a
phase shift r\ in the voltage wave dV. By a
proper selection of antenna coupling, it is often
possible to operate near a resonance of the
driving circuit, whereupon Tp approaches zero
and there is little or no phase shift q.

Inasmuch as dV is proportional to M0, a
space plot of the variation of antenna imped-
ance will likewise be a space plot of the varia-
tion of output voltage. This is a most impor-
tant point to remember. For it means that the
voltage out of the r-f system can be plotted
point by point for a slow relative motion of
fuze and target to give detailed information
on the performance in rapid motion. All that
need be changed is the time scale ; the fuze an-
tenna goes through its sequence of variations in
whatever time is required for the fuze to move
through the region of influence, and the wave
form will be identical in every case. This as-

ANTENNA IMPEDANCE

sumes, of course, that the circuit is capable of
following the time variations, which experi-
ence has shown is no restriction.

The space variation of M which gives the
voltage variation has been called the M wave
and is so referred to in the discussion which
follows. Extensive use is made of point-by-point
plots of the M wave in testing fuzes.

We note that S may be measured as dictated
by convenience in terms of either series or
parallel components, and that the complex form
of S is completely specified by either set of
measurements, as shown by equations (89),
(90), and (91). In many cases Ts/QSs and
Tp/QSp are small compared with unity, so that

Ss = — Sp = zSzSq = ±|*S|.

In any case the basic equation (89) which
represents the situation is

dV = MS = M0S0eK«-'K

We apply this to the two special cases in which
we are interested.

For the ground-approach case we have, uti-
lizing equation (45a),

M = Si, G^e^e’a- (03)

For the airborne target we have from equa-
tion (50) #

M = (cos r)e/a. (94)

Subscripts previously used in connection with
A, G, /, 0, and $ will no longer be carried ex-
cept in cases where there may exist a possibility
of misunderstanding.

2 7 ANTENNA IMPEDANCE0

The previous section has shown that a knowl-
edge of the components of Zxu the free-space
antenna impedance, is essential if circuit re-
sponse to the reflected impedance arising from
reflection is to be predicted. This section is con-
cerned with the values of antenna resistance
and reactance observed in actual cases.

c The following bibliographical references are perti-
nent to this section: 6, 8, 11, 28, 30-34, 37, 38, 49, 50, 52,
54-60, 62-67, 69.

Specifically the resistance and reactance sen-
sitivities of the fuze circuit are functions of
( XS,RJ or (XV,RV) and must be evaluated at
the particular operating point characteristic of
the particular antenna used. Since the various
missile-antenna combinations present widely
different impedances, it becomes necessary to
measure, or in some cases to calculate, the sen-
sitivity parameters for a given circuit over a
large range of load impedance, so that the an-
tenna can be designed to have an operating
point as near as possible to the optimum point
for circuit response. It should be pointed out
here that there are limitations to the antenna
impedance that can be achieved within the
limits set by the tactical situation and by speci-
fied military characteristics. Likewise the val-
ues of S can be changed by circuit design only,
within certain limits set by present-day vacuum
tubes. Thus antenna design must be coordi-
nated with circuit design to give optimum per-
formance within the limits of both. In addition
it is necessary to set up dummy antennas for
testing fuzes. A knowledge of actual antenna
impedance is essential here also. In this section
we are concerned with the antenna impedances
that can be effectively achieved. Chapter 3 will
give details about the circuit performance
under the load and load variations presented
by the antenna.

2-71 Specification of Antenna Terminals

In all the previous discussion the fuze an-
tenna has been treated as a two-terminal box
which sends out radiation. No detailed knowl-
edge of the internal circuit was assumed. We
imagine this antenna to be connected to a
circuit whose sensitivity is specified. Now when
the whole is connected together, a given reflec-
tor in space sets up a certain A V in the r-f
circuit. Once the whole arrangement is con-
nected, the antenna terminals lose their iden-
tity and their location becomes arbitrary. Thus
if we open the arrangement at any two points
(see Figure 9) and call everything on one side
of the cut the fuze circuit and everything on
the other side the antenna, the values of Xv, Rp
and Sp and T}) must be so related that the value

SECRET

THE RADIATION INTERACTION SYSTEM

of AF calculated by their use is independent of
the particular pair of points selected as an-
tenna terminals.

Thus in specifying fuze performance or an-
tenna performance we are at liberty to select
any two terminals within the network as an-
tenna terminals. It has been customary in vari-
able-time [VT] fuze work to consider all the
circuit elements inside the fuze electronic as
the fuze circuit, even if the assembly contained
some antenna impedance matching network,
and consider the points where the leads from
this circuit are connected to the external radi-

III

□

TARGET

Figure 9. Arbitrary division into fuze circuit
and antenna.

ating system as the antenna terminals. The dis-
cussion of the antenna problem has assumed
that the only ohmic losses in the antenna are
radiation losses. Experiments have shown this
to be a valid assumption with the antenna ter-
minals just specified. If some other pair of
points is selected so that some energy-absorb-
ing coupling elements are included in the net-
work, due account must be taken of these losses.

2 7 2 Experimental Measurement of RP

The parallel radiation resistance Rp is meas-
ured by a substitution method. A typical fuze
circuit is used as an indicator. In the fuze
circuit several quantities, such as diode voltage
V d or oscillator grid voltage Eg, oscillator plate
current Ip, and carrier frequency /, are all func-
tions of (Xp,Rp). The fuze is first placed on the
projectile on a high stand and the free-space
values Vd or Eg, Ip, and / noted. The fuze is
then removed from the projectile and placed in
a shield box. Reactance and resistance are
added to the antenna terminals until the free-
space values are duplicated. It is assumed that
the shield box does not load the fuze at all, so
that the amount of resistance across the ter-

minals when the load is duplicated repre-
sents Rp.

In making this measurement the exciting cap
or bars are often not removed. The resistors
are merely substituted across the points that
have been previously agreed upon as the an-
tenna terminals. The shield adds reactance to
the antenna so that direct measurements of Xp
are not obtained this way.

To show that the shield does not introduce
serious losses, the whole antenna is removed
from the fuze and resistors substituted directly
across the fuze terminals. The size of the ar-
rangement is so small compared to X that radia-
tion is negligible and the substituted resistance
represents Rp . By such tests it has been shown
that shield losses are negligible, so that the
more convenient shield box can be used where
desired. In making this comparison test it is
necessary to remove dielectric insulators so
that losses in them do not confuse the measure-
ments.

In the case of fuzes which use the projectile
as the antenna, the radiation load is removed
by putting a shield can over the nose fuze, as
shown in Figure 10. Such an arrangement re-
places radiating currents in the antenna by
nonradiating currents inside the shield can and
thus substitutes can losses for antenna losses
in measuring Rp. If both are small compared
with the true radiation losses, this substitution
causes negligible error. The final proof that the
errors are negligible is obtained by making a
pole test (see Section 2.12) and comparing
actual signal with calculated signal based upon

SHIELD BOX

Figure 10. Method of removing radiation load
without removing projectile.

the measured values of Rp and Sp. Within an ex-
perimental error of less than 10 per cent there
is agreement.

It is interesting to note here that the absolute
ohmic value of the substitution resistors used
for the measurement of Rp need not be known.

SECRET

ANTENNA IMPEDANCE

Since M is proportional to dRp/Rp, only ratios
of resistances are needed, and any set of resis-
tors whose r-f resistance is a constant fraction
of the d-c resistance may be used. It has been
found that International Resistance Company
[IRC] type F-l or F-% resistors fulfill this
requirement; in fact their r-f values are quite
close to their d-c values. However, if it is of
importance to know the power radiated, as
is the case in jamming calculations, the true
value of the r-f resistance must be known. It
has been customary for all collaborating lab-
oratories to use equivalent sets of r-f resistors
supplied by a central laboratory.

2 7-3 Specification of Xp

The quantity Tp/QSp appearing in equation
(91) is in many cases so small that it can be
neglected, as has already been mentioned. To
see this, the value of TP/QSP at the operating
point must be evaluated, a procedure which re-
quires knowledge of Xp. The question immedi-
ately arises: Is the total apparent reactance
across the antenna terminals the correct value
of Xp, or should we use only the part that
appears by virtue of radiation? The argument
above about specification of antenna terminals
implies that either arrangement should give
the same answer. The r-f section of the fuze
can be represented in block diagram as in Fig-
ure 9. The target in space gives rise to a cer-
tain XV out of the terminals to the audio-con-
trol circuit. We are at liberty to divide the
whole arrangement at any convenient place by
a line xx and call the left part the fuze and
the right part the antenna. If the calculations
are performed properly, the result must be the
same no matter where xx is chosen. We choose
to put the fixed part of the antenna reactance
to the right of xx and call it a part of the an-
tenna. Similarly, if there are dielectric losses
in the antenna mounting, we can assume them
to be represented as a resistance across the
antenna terminals and put it to the left of xx
as a part of the fuze circuit. This is the adopted
convention. As a matter of fact we can, if we
desire, divide the Xp any way we choose be-
tween fuze circuit and antenna.

To illustrate the point we show that the quan-
tity Tp/Q is independent of how Xp is defined

Suppose

dXp’

(95)

This does not
Then

2 xfC9

affect the generality of the result.

dV = dV dCp
dX p dC p dX p

and

Tp = XI dV_ = -1

Q Rp dXp fRp dC p

(96)

(97)

which shows that Tp/Q is independent of how
Cp is defined.

2 7,4 Measurement of Xp

In the following discussion Xp is considered
as the total reactance across the antenna ter-
minals. The term Xp is measured by a direct
substitution method. The fuze is mounted on a
missile, and values of Vd or Eg, Ip, and / are
recorded. The fuze is then removed from the
vehicle and the circuit disconnected from the
antenna at the previously selected terminals.
Resistors and condensers are placed across the
terminals until Vd, Ip, Eg , and / are duplicated,
thus duplicating Rp and Xp. For all fuze designs
now used Xp is capacitative. To a good approxi-
mation Xp is the capacity across the antenna
terminals as measured by low-frequency meth-
ods. This is shown by the fact that Xp varies
only slightly with projectile size.31

2.7.5 Effect of Feed Geometry

upon Rp and Xp

Figure 11 shows the effect upon Rp of chang-
ing the size of the exciting ring on the T-50
type of fuze, with the spacing from ring to
ground held constant at 1 in. Results are shown
for several bombs, two carrier frequencies
(White and Brown) for ring lengths ranging

THE RADIATION INTERACTION SYSTEM

from 14 to 3 in. As expected, an increase in
ring length decreases Rp.

The effect of ring length upon Cp is shown in
Figures 12 and 13 for Brown and White fre-

Figure 11. Rp as function of ring size; BRLG-
type fuze; gap width, 1 in.; solid lines, Brown
frequency; broken lines, White frequency; curves
1, M-30 100-lb bomb; curves 3, M-64 500-lb bomb;
curves 4, M-65 1,000-lb bomb; curves 5, M-66
2,000-lb bomb; curves 6, M-81 260-lb bomb.

quencies respectively, with a constant gap size
of 1 in. In Figures 12 and 13 the capacity is
shown as 6.9 ppf for all the bombs, for a ring
length of 1 in. This is not precisely correct, as
capacities associated with the various projec-
tiles vary somewhat; the curves are meant to
indicate the variation of the capacity with ring
size. The value 6.9 ppf is not far from correct,
however; for all the bombs, the true range of
values at the 1-in. point is about 6.9 ± 0.5 qpf.
An increase in ring length increases Cp.

The effect of gap size upon Rp and Cv is
shown in Figure 14. The size of the gap in the
range shown (spacings from 1/2 to 1 in.) has

virtually no effect upon Rp. The term Cp, of
course, decreases as the spacing is increased.
It thus becomes possible, within the limitations
of space requirements, to vary the gap size to
bring Cp to a favorable value without affect-
ing Rp.

In the case of transverse center-fed dipoles
(T-51, T-82) Cp is increased by increasing the
size of the dipoles or by reducing their separa-
tion. The term Rp decreases when the size of
the dipole is increased While it is an

advantage to get lower Rp by using longer
dipoles, air resistance and operational difficul-

1/4 1/21 2 3

RING SIZE (INCHES)

Figure 12. CP as function of ring size; BRLG-
type fuze; gap width, 1-in.; Brown frequency;
curve 1, M-30 100-lb bomb; curve 3, M-64 500-lb
bomb; curve 4, M-65 1,000-lb bomb; curve 5,
M-66 2,000-lb bomb; curve 6, M-81 260-lb bomb.

ties increase with increasing length, and these
considerations serve to limit the length of an-
tenna which may be used. Electric efficiency
must be subordinated in this design.

SECRET

ANTENNA IMPEDANCE

Typical Values of Rp and Xp

Figure 15 shows the value of Rp as a function
of carrier frequency for several common bombs
using a standard T-50 ring; the feed must be
specified since Rp depends on it. The large
range of values for all bombs at any one fre-

RING SIZE (INCHES)

Figure 13. CP as function of ring size; BRLG-
type fuze; gap width, 1 in.; White frequency;
curve 1, M-30 100-lb bomb; curve 3, M-64 500-
lb bomb; curve 4, M-65 1,000-lb bomb; curve 5,
M-66 2,000-lb bomb; curve 6, M-81 260-lb bomb.

quency illustrates clearly the difficulty of de-
signing a single fuze which will work on all
bombs. It was for this reason among others
that the T-51 type transverse antenna fuze was
designed. The T-51 with its independent an-
tenna has radiation resistance relatively inde-
pendent of projectile size.

The logarithmic spread in Rp among the
vehicles tends to decrease as the frequency is
lowered, The mean value of Rp increases. Until
recently, as shown in the next chapter, it was

1/2 5/8 3/4 7/8 I

S PACING - RING TO BASE OF CAP (INCHES)

Figure 14. Effect of gap size upon RP and CP ;
BRLG-type fuze; M-65 1,000-lb bomb; White
frequency.

not feasible to operate fuze circuits into a mean
value of Rp as high as 40,000 ohms. The im-
proved reaction grid detector [RGD] circuit

B-35 BH5 B+5 W-IO W+IO W+30

FREQUENCY

Figure 15. RP as function of carrier frequency;
T-50 ring; curve 1, M-30 100-lb bomb; curve 2,
M-57 250-lb bomb; curve 3, M-64 500-lb bomb;
curve 4, M-65 1,000-lb bomb; curve 5, M-66 2,000-
lb bomb; curve 6, M-81 260-lb bomb.

(SWHOI

THE RADIATION INTERACTION SYSTEM

5" HVAR

Figure 16. Scale outline drawings of a number of missiles for which VT fuzes were designed. Fuzes
with longitudinal excitation were designed for all missiles shown. Transverse antenna fuzes were also
designed for bombs. To show the relative size of fuze and missile, the outline of the fuze is shaded.

SECRET

DIRECTIVITY PATTERNS

described in Section 3.1 makes it possible to
use one frequency for a large number of projec-
tiles by allowing operation at a high value
of Rp.

Complete tables of Rp and Xp are not avail-
able for all projectiles. Table 1 gives the values
at important frequencies for fuze projectile
combinations of current interest. The range of
missile sizes for which fuzes were designed is
illustrated in Figure 16. Photographs of typical
fuze and missile combinations are shown in
Figure 7 of Chapter 1.

Table 1. Typical values of Rp and Xp for various
fuze-projectile combinations.

Fuze

type

Projectile

Carrier

frequency

(ohms,

ap-

prox.)

(ohms,

ap-

prox.)

T-91

M-30 bomb
(100-lb GP)

Brown

20,000

300

T-91

M-66 bomb
(2,000-lb GP)

Brown

40,000

300

T-92

M-64 bomb
(500-lb GP)

White

11,000

200

T-92

M-65 bomb
(1,000-lb GP)

White

10,000

200

T-132

M-43 mortar
with M-56 tail

White + 20

20,000

150

T-132

M-56 mortar

White + 20

6,000

150

T-171

M-43 mortar
with M-56 tail

Brown

90,000

300

T-171

M-56 mortar

Brown

60,000

150

M-166

White + 35

150,000

600

T-2005

HVAR rocket

Brown

3,800

150

AR 5-in. rocket

Brown

3,600

150

T-5, T-6

M-8,4|-in.

rocket

White + 30

8,000

300

28 DIRECTIVITY PATTERNSd

We come now to a more detailed study of the
properties of /2(0,<£), the power radiation
pattern or directivity pattern. The importance
of /2(0,0 ) was indicated in Section 2.5 above,
where it was shown that the reflected impedance
Zr, due to an object situated in a direction
(0i3><£i3) relative to the fuze, is proportional to
fi (013,013). Furthermore, the gain G is a func-
tion of

d Bibliographical references pertinent to this section
are 6, 8, 36, 40, 47, 50, 52, 54-60, 62, 63, 64, 68.

281 Measurement of Directivity Patterns
Experimental Setup

The simple convenient method which has
been devised for the measurement of directivity
patterns may be understood with the help of
the photographs in Figures 17, 18, and 19. In
Figure 17 the antenna, which consists in this
case of the projectile plus the fuze in its nose,
is mounted horizontally on a platform about
15 ft above the ground. The platform is free
to rotate about a vertical axis (Figures 18 and
19). The receiver (Figure 17), consisting of a
dipole antenna feeding a detector, is situated
about 150 ft from the transmitter. Power is fed
to the transmitting antenna by means of a care-
fully choked line coming from the power supply
on the ground. The line is attached to the an-
tenna at a voltage node. The whole setup is
situated in an open field. The radiating antenna
can be rotated through any angle and the re-
ceiver signal plotted as a function of this angle.
The plate supply of the transmitting oscillator
is based upon an a-c source of 60 c. Full-wave
rectification with no filtering is used; there is
thus a plate modulation frequency of 120 c
which can be detected at the receiver.

In the type of fuze antenna shown in the
photograph, the directivity pattern has cylin-
drical symmetry because of the symmetry of
the projectile. In such a case the directivity has
no dependence upon 0, and the pattern may be
represented analytically as /2(0). The angle
of rotation about a vertical axis is then equal to
6 ; when the nose points directly at the receiver
^ = 0°. The detector used is of the square law
variety, so that the audio signal is directly pro-
portional to the square of the field strength or
to f2(0).

Where cylindrical symmetry is not present,
the pattern must be taken in several planes to
get a reasonably complete set of values for
f2 (0,0) . Such asymmetry may occur when the
antenna is separate from the bomb, as in the
T-51 and T-82 type designs, the bomb acting as
a parasitic reflector or director.

Detector Circuit

Some notes may be added concerning the de-

SECRET

:ET|

THE RADIATION INTERACTION SYSTEM

tector circuit, shown in Figure 20. The circuit
and physical layout are symmetrical, with a
view to obtaining balance with respect to
ground. This has the effect of minimizing the
effect of any vertically polarized components of
the field caused by reflection from the ground,
which otherwise could alter the apparent shape

Figure 17. Field setup for obtaining directivity
patterns.

of the directivity pattern, especially affecting
the symmetry of the patterns.

The r-f chokes are for the purpose of reduc-
ing interference from transmitters operating
in or around the broadcast region. These chokes
have a high impedance at the carrier frequen-
cies used in fuze antennas but a low impedance
at lower frequencies. The coupling condensers

vacuum-tube voltmeter, preceded if necessary
by an amplifier.

The output of the detector follows quite accu-
rately a square law for outputs up to 100 mv.

Figure 19. Platform of Figure 18 shown
lowered to ground.

The output may be kept below this level by
adjusting the power supply feeding the trans-
mitting antenna.

Figure 18. Rotatable platform holding fuzed
bomb for directivity pattern measurements.

Reflections from Ground

The reflections from the ground contribute
to the output of the detector and therefore may

in the grid circuits prevent the grids from
being shorted by the chokes at direct current.
The condenser from the plates to ground serves
as an r-f by-pass; it has a high impedance to
audio frequencies.

The audio output is fed to a Ballantine-type

Figure 20. Detector circuit; components are
enclosed in metal box mounted between arms of
receiving dipole.

change the apparent directivity pattern. This
matter has been studied and is presented in
some detail as supplementary material in Sec-

DIRECTIVITY PATTERNS

tion 2.15, where it is shown that the ground
reflection introduces negligible error for the
simple radiation patterns now used.

By means of the equipment described above
a large number of directivity patterns have
been obtained. The patterns fall into two
classes: (1) patterns for fuzes which use the
projectile as the antenna (longitudinal excita-
tion), and (2) patterns for fuzes which use a
separate antenna such as a short transverse

Figure 21. Directivity pattern for M-64 bomb
at B — 16; longitudinal excitation; G = 1.55.

dipole or loop (transverse excitation). The pat-
terns will be discussed according to this classi-
fication.

Longitudinal Excitation
Typical Patterns

In longitudinal excitation the fuze proper is
connected to the projectile at one end. The an-
tenna, consisting of fuze and projectile, is split
by an insulator near the end for the purpose
of feeding energy to it. The exact position and
size of the gap over the range used, while im-
portant in determining the antenna impedance,
have little effect upon the pattern. Therefore it
will not be necessary to specify the feed exactly

in describing the patterns for longitudinal ex-
citation.

A preliminary idea of the character of the
patterns may be obtained from Figures 21, 22,
and 23. These show the directivity patterns
f2(6) plotted versus 0 on polar coordinate
paper; for these antennas, the directivity pat-
tern is not a function of <£, cylindrical symme-
try being present. Figures 21, 22, and 23, for
a 500-lb GP bomb (M-64) at three carrier fre-

Figure 22. Directivity pattern for M-64 bomb
at B -f- 15; longitudinal excitation; G = 1.9.

quencies, demonstrate the effect of changing
the frequency. As the carrier frequency is
raised, which means the antenna becomes elec-
trically longer, the pattern departs more and
more from the simple sin 6 pattern of an ele-
mentary dipole (shown in Figure 24). The
minor lobes in the pattern for a carrier fre-
quency of W + 10 (see Figure 23), will be
noted. In these patterns, as in all the patterns
obtained with longitudinal excitation, the radi-
ation is more intense off the end of the antenna
away from the feed point. This will be dis-
cussed below in greater detail. The values of G
for each pattern are given in the captions to the
figures. These values were obtained by graphi-
cal integration of the patterns. The effect of
using one frequency for exciting various pro-

THE RADIATION INTERACTION SYSTEM

jectiles is shown in Figure 25. Figure 25 repre-
sents a series of patterns at W + 10 for the
M-30, M-81, M-64, M-65, and M-66 bombs. In
this figure the patterns are plotted in rectangu-
lar coordinates.

Since tactical utility has required that the
fuze be located in the nose of the projectile, we
are interested in the values of f2(0,<f>) in front
of the equatorial plane, i.e., for 6 < 90 degrees
in the patterns of Figure 25. It is immediately

Figure 23. Directivity pattern for M-64 bomb
at W + 10; longitudinal excitation; G = 2.6.

may be expected from qualitative arguments.
The antenna may be thought of, crudely, as a
piece of transmission line with a generator at
one end and an impedance at the other end. A
wave starts out from the generator ; some of it
is absorbed in the impedance at the other end
(radiation) ; the rest is reflected back. Since
the amplitude of the wave traveling from the
generator is greater than that of the return
wave, the part of the radiation due to the for-

Figure 24. Directivity pattern for infinitesimal
dipole; /2(0) = sin2(0) ; G — 1.5.

evident that there is a wide range in Zr for
targets in the range 10 to 90 degrees off the
nose. This is a complicating factor which re-
quires that more than one frequency be used
in designing longitudinal antenna fuzes.

This is an unfortunate complication that
could largely be avoided if the feedpoint could
be located in the rear of the projectile.

General Features of Longitudinal
Patterns

The directivity patterns obtained with longi-
tudinal excitation have several general fea-
tures worthy of note. Some of these have
already been mentioned and will be treated
here in somewhat more detail.

“Lean” of Patterns. The patterns “lean”
away from the feedpoint. This characteristic

ward wave, primarily forward radiation, is
more dominant than the part due to the return
wave.

For end-fed antennas, at a given frequency,
the asymmetry is greater the greater the thick-
ness of the antenna relative to its length. Cen-
ter-fed antennas, even if of considerable thick-
ness, have symmetrical patterns.

Patterns like those experimentally obtained
may be computed by assuming an antenna cur-
rent distribution with features as above de-
scribed. That is, suppose we assume that the
antenna current I is given by an expression of
the form :

/ = ejUt - z) _ RIie-j(2n/\)(L - z)

(98)

In equation (98), h represents the amplitude

SECRET

DIRECTIVITY PATTERNS

of a wave traveling in the positive z direction,
z is the running coordinate of the antenna with
the feedpoint at the end z = 0, L is the length
of the antenna, and RIX represents the ampli-
tude of a return wave. Thus R is a reflection
coefficient whose magnitude is less than unity,
and 5 represents a phase shift occurring at re-
flection. Thus 1 represents a return wave super-
posed upon a forward wave.

Now the angular dependence E (6) of the

//}

'il

f¥

' H

\ III

- 1
//

# \

\ HI

v//

. 0

/ • A 1

1 ^

./ .j—

a n 1 AA 1

or\ IAA u

in ion

e (DEGREES)

Figure 25. Directivity patterns at W + 10;
longitudinal excitation; curve 1, M-30 100-lb
bomb; curve 3, M-64 500-lb bomb; curve 4, M-64
1,000-lb bomb; curve 5, M-66 2,000-lb bomb;
curve 6, M-81 260-lb bomb.

remote radiation field produced by a linear an-
tenna of length L is given by

E(e) a sin 9 J I(z)e l * cos e\ yZ} (99)

where I (z) is the current distribution along z.

It has been found that if I (z) be taken as in
equation (98), then normalized values of E2(6)
obtained from equation (99) agree well with
experimentally obtained patterns when R and 5
are adjusted empirically. The picture of a for-
ward wave and a return wave appears to be
adequately correct to allow extrapolation and
interpolation for changes in antenna size.

Small-Angle Radiation. Experimental meas-
urement of patterns and the theoretical compu-
tations of patterns outlined above (see Section
2.10) both lead to the conclusion that, for small
angles, we may express the directivity pattern
as

P(6) = a sin2 d, (100)

where a is a different constant for each projec-
tile. This is a valuable generalization that facil-
itates computation of Zr for cases arising in
practice.

Effect of Projectile Geometry. Because of the
considerable thickness of the projectile anten-
nas, their physical lengths are considerably less
than their “electrical” lengths. For a given
physical length, an increase in thickness serves
to increase the electrical length.

Effect of Tuning or Loading. The pattern de-
pends only upon the frequency and the antenna
geometry. There is, of course, no effect due to
tuning or loading the antenna circuit.

Comparison of Patterns for Fuze Work

One of the desiderata of a proximity fuze is
that it be usable without modification on vari-
ous projectiles. It thus becomes necessary to
examine, among other things, the variations in
the directivity patterns for the various projec-
tiles. This variation, for longitudinal excitation,
has already been illustrated in Figure 25, where
it is shown that at a particular frequency the
patterns vary considerably. In the search for
an optimum operating frequency for a prox-
imity fuze for bombs, a large number of direc-
tivity patterns was taken at various frequencies
for the several bombs and a comparison of
f2(0) was made. In this comparison, relatively
small values of 0 are of importance in the
ground-approach application since it is these
that are encountered under terminal conditions
for ordinary bomb releases. Because of the ap-
proximate law f2{6) — a sin2 6, relative values
for the various projectiles at one angle, i.e.,
0 =: 30 degrees, will hold, roughly, for smaller
angles. Figure 26 presents the values of

for various bombs for a range of frequencies.
From Figure 26 it is seen that the expected

SECRET \

THE RADIATION INTERACTION SYSTEM

variation in signal covers a very wide range,
except at frequencies of 50 me and below. This
agrees with the discussion in Section 2.5.6,
which indicated that the signals for various
bombs tend to approach the same level as the
electrical length of the antenna is shortened
below 1/2. Until close to the end of World
War II it was not feasible to use frequencies
as low as 50 me, because the values of parallel
radiation resistance encountered at these low
frequencies (see Figure 15) would not permit
efficient matching to the driving circuit, thus

Figure 26. Relative signal strength Mo/h =
(X/47 r) Gp(e) at 6 = 30°, over a range of fre-
quencies; longitudinal excitation; curve 1, M-30
100-lb bomb; curve 2, M-57 250-lb bomb; curve
3, M-64 500-lb bomb; curve 4, M-65 1,000-lb
bomb; curve 5, M-66 2,000-lb bomb.

leading to low S values for high Rp. If the re-
flected signal is to be nearly the same at all
heights for different projectiles using the same
fuze, the product S0M0 must be constant. The
term M0 depends upon the directivity pattern
and So depends upon the operating point Rp.
To hold the spread of S0M0 to reasonable values,
it was found necessary to use two different
radio frequencies in order to accommodate the
fuze to various bomb sizes. It will be seen in
Section 2.8.4. that transverse excitation largely
avoids this difficulty. Toward the end of World
War II a circuit was developed which could be
matched to the high values of parallel radiation

resistance encountered at low carrier frequen-
cies. This made possible the design of a more
nearly universal longitudinal excitation fuze for
bombs.40

Transverse Excitation
Transverse Dipole

Fuzes working on this principle use a short
transverse dipole as an antenna. The body of
the projectile is not intentionally used as a part
of the fuze, although it introduces complica-
tions as shown in Section 2.5.4. Space limita-
tions are such that the transverse dipole is
short compared to a half wave, and the direc-
tivity pattern tends to be like that for a short
thin wire antenna, f2{6) — sin2 6 (see Figure
24). The close presence of the body of the pro-
jectile modifies the pattern so that it is no
longer a figure of revolution about the antenna
axis. In some cases the projectile acts like a
director, making the radiation toward the back
of the projectile greater than toward the front.
A certain amount of asymmetry with respect
to the bomb axis is sometimes present because
of a slight unbalance in the feed.

An unbalanced feed for the transverse dipole,
aside from the effects discussed in Section
2.5.4, gives rise to a directivity pattern which
does not have axial symmetry. In Section 2.5.4
it was seen that the longitudinal currents give
rise to a correction which is small if the longi-
tudinal currents are kept small.

To verify the fact that these currents are
small the radiation pattern of the fuze-projec-
tile combination is measured with equipment
arranged similarly to that shown in Figure 17.
The projectile axis is horizontal and the axis
of the transverse dipole is vertical. The re-
ceiver, which is sensitive only to horizontally
polarized radiation, does not receive the energy
radiated by the transverse currents flowing in
the fuze dipole and the projectile behind it. It
receives only the radiation from the longitudi-
nal currents and gives a pattern like those for
axial feed (see Figures 21, 22, 23, and 25).

The strength of the axial radiation is com-
pared with the strength of the dipole radiation
by putting the fuze dipole in the horizontal posi-
tion and pointing the projectile directly toward

SECRET

DIRECTIVITY PATTERNS

or away from the receiver. In this orientation
the longitudinal currents do not radiate toward
the receiver, and the received signal is that
from the transverse dipole alone, modified by
the reflecting properties of the projectile. As a
result of the two measurements, two field in-
tensities are obtained which show the relative
amounts of energy radiated by the longitudinal
and transverse currents. In order to suppress
the longitudinal currents it has been found
necessary to use a relatively long wavelength
so that the projectile is nonresonant.

When the directivity pattern is measured with
fuze dipole horizontal and projectile horizontal,
an asymmetric pattern like that in Figures 27

Figure 27. Directivity pattern for M-57 bomb
at W -(- 35; transverse excitation; pattern taken
in plane determined by longitudinal axes of dipole
and bomb.

and 28 is obtained. The right-left asymmetry
arises from the addition of the patterns from
longitudinal and transverse currents. The fore-
aft asymmetry arises from the reflecting prop-
erties of the projectile. For short projectiles
(see Figure 27) the fore-aft asymmetry is less
marked than for larger ones.

As we have seen, if the longitudinal current
is small its effects can be neglected. Thus, to a
good working approximation, we can take the
directivity in the directly forward direction to
be unity and the directivity in other directions

forward of the equatorial plane as cos2 a or
sin2 6.

When the transverse dipole is used, the size
of the projectile has but little effect on the radia-
tion resistance, provided the diameter is not too

Figure 28. Directivity pattern for M-64 bomb
at W + 35; transverse excitation; pattern taken
in plane determined by longitudinal axes of
dipole and bomb.

large and the projectile does not form an effec-
tive shield by imaging the dipole. This type of
antenna is effective on all types of American
bombs which the fuze will fit. It is not satisfac-
tory with the British-type square-nosed bombs
like the 4,000-lb LC, since this kind of bomb
forms an effective shield unless the fuze is
mounted on an extension. The British have
found by extensive tests that a short extension
makes the fuze quite satisfactory on this bomb.

Furthermore, the vehicle has only a relatively
minor effect on the pattern, as we have seen, so
that the fuze operation becomes relatively inde-
pendent of the size of the projectile. In addi-
tion, the transverse type of excitation will op-
erate with nonconducting projectiles, such as
the plywood belly tanks arranged with bomb
fins used as fire bombs.

Loop Excitation

It is possible to obtain transverse excitation
by means of a transverse magnetic dipole. This

THE RADIATION INTERACTION SYSTEM

is achieved by means of a small loop antenna,
about 3 in. in diameter, whose plane includes
the axis of the projectile, as in the T-172 fuze.
The polarization of the radiation is different
from that of an electric dipole, as shown by
Table 2.

Table 2. Polarization of radiation in loop and dipole fuzes.

Dipole

Loop

— cos a cos 5

A . 9
— sin 5

A .

— sin 5

— cos a cos 5

The coordinate system for Table 2 is as de-
fined in Section 2.5.4. Aside from the polariza-
tion change the argument is similar to that
outlined for the transverse dipole, including
unbalance effects. Similar radiation measure-
ments are required, with due consideration for
the polarization.

29 WORKING SIGNALS;

GROUND-APPROACH CASEe

This chapter is primarily concerned with the
variations of antenna impedance as the fuze
approaches a reflecting target. This variation
has been described by M, and the variation of M
from point to point in space has been called the
M wave. It has also been shown that the voltage
change dV out of the r-f system can be specified
in terms of circuit parameters.

dV = MS. (89)

In other words, the voltage out of the r-f system
is proportional to M, and in so far as relative
wave form and amplitude are concerned M can
be considered as a working signal set up by the
reflecting target. Chapter 3 deals with the prop-
erties of circuits and the values of S that can
be achieved.

The method of utilization of this signal has
been indicated in Chapter 1 and will be briefly
recapitulated here. The voltage dV is applied
to an amplifier; the output of the amplifier is
applied to the grid of a thyratron; when the

output of the amplifier is of the proper magni-
tude and phase the thyratron discharges
through a detonator which initiates the firing
train resulting in the burst. In most cases the
transmission time of the signal through the
fuze and detonator are negligible. A treatment
of the delay to be expected is given in Chapter
3, which deals with the audio amplifier and
firing circuit. Since delays are generally small,
the basic problem of the design is to make the
output voltage of the amplifier reach the firing
level at the moment when the projectile is in a
position such that its burst would do the maxi-
mum amount of damage. While the necessary
adjustments could be made empirically upon
the basis of field trials, it is extremely helpful
to be able to predict the expected point of func-
tion from the fuze parameters and the ballistic
problem. Such a knowledge allows treatment
of many cases based upon performance in a
typical case; it also aids recognition of abnor-
mal performance.

We proceed to show how the prediction is
made for the case of a bomb approaching the
ground. For the sake of clarity we first treat a
special case in Section 2.9.1 and then turn to
a discussion of each of the factors involved in
Section 2.9.2.

2 91 Prediction of Height of Function

The case selected is that of the ring-type fuze
(longitudinal excitation and White frequency
band) on the M-64 (500-lb) bomb, released
from level flight by an airplane flying at 200
mph from an altitude of 10,000 ft over earth
which has a reflection coefficient of 0.5.

The equation governing this case, equation
(93), has been derived in Section 2.6.2. Utiliz-
ing it, we have

dV = MS = G/2(0,0)ey[(-4'vx) + J

(101)

Equation (102) represents an audio-frequency
voltage with peak amplitude

MoS0 = (102)

and frequency

." *

e Bibliographical references pertinent to this section
are 12, 16-22, 27, 35, 39, 40, 70, 71, 74.

SECRET

WORKING SIGNALS; GROUND-APPROACH CASE

For a falling bomb dh/dt is essentially con-
stant over the last few hundred feet of flight,
and we can take dh/ dt as the vertical component
of the striking velocity. Thus equation (101)
represents a voltage of constant frequency and
rising amplitude in the range in which there
is appreciable reflected signal.

Let us assume that the steady-state voltage
amplification of the amplifier is known in the
form of a curve g(F), henceforth denoted sim-
ply by g and that the net holding bias of the
firing thyratron is B. Then we can say that the
height of burst h is approximately

h = XnSo £ r-(e,<t>) (104)

Equation (104) is based upon the tacit as-
sumption that there is no delay in the amplifier
and detonator, and further it ignores the fact
that the thyratron can only fire when the volt-
age is positive, thereby introducing an uncer-
tainty in the height of operation of approxi-
mately (X/ 2). The nature of these corrections
will be discussed in more detail in Section 2.9.2.

The quantity B/g represents the peak voltage
into the audio-control circuit that is necessary
to fire the detonator.

We now insert appropriate values in equation
(104) for our special case as follows: G/4jc =

0.208, the vertical component of striking veloc-
ity = 740 fps, and the striking angle = 18.5 de-
grees, the wavelength = 8.2 ft, and /2(18.5 de-
grees) = 0.085. The audio frequency F is
(2 X 740) /8.2 = 180 c. Typical values of G, B ,
and So, are 80, 4.4, and 15, respectively. Using
these values, equation (104) gives h — 20 ft.

This is the height of burst to be expected if
only radiation fields are involved. Actually the
induction field introduces a correction, as shown
in Section 2.10, when the magnitude of h ap-
proaches X.

2,9,2 Factors Affecting Magnitude and
Frequency of Impedance Signal M

Having taken a brief overall view of the vari-
ous factors determining the point of function
of a fuze by computing the position of the burst
in a typical case, we shall now discuss in more

detail the various factors affecting the im-
pedance signal M. We use the term “signal”
advisedly, because the impedance change is de-
pendent upon the interaction between fuze an-
tenna and target, and therefore contains intelli-
gence as to the conditions of such interaction.

The amplitude and frequency of the imped-
ance signal or M wave depend upon a variety
of conditions, some of which depend upon the
fuze design and the projectile on which the
fuze is used, others depending on ballistics and
reflecting properties of the target. We shall
omit factors depending upon the circuit adjust-
ment, which is the subject of the next chapter.
Aside from these the factors affecting M may
be grouped according to the following scheme.

Fuze antenna factors

1. Directivity pattern.

2. Antenna gain.

3. Carrier frequency.

Ballistic and target factors

1. Distance from target.

2. Orientation of fuze antenna relative to
target.

3. Speed of approach to target.

4. Reflecting properties of target.

We shall discuss each of these factors in turn.

Fuze Antenna Factors

Directivity Pattern. It has been shown, equa-
tion (93), that M is proportional to f2(6,(f>);
the values of 6 and </> in question are those ob-
tained by drawing a straight line from the an-
tenna perpendicular to the ground.

The nature of the patterns in use has been
discussed in Section 2.8, where there was also
presented a series of typical patterns. Upon
referring to these patterns, it is evident that
in the case of longitudinal excitation f2(0) is
relatively small when the projectile is vertical
or nearly vertical to ground, and becomes larger
as the projectile becomes more nearly parallel
to the ground.

In the case of transverse excitation, on the
other hand, /2(0,<£) is large when the projectile
is normal to the target surface. When the angle
between the surface and the axis of the
projectile is any value other than 90 degrees,

■

^SECRET

THE RADIATION INTERACTION SYSTEM

there is an uncertainty in the value of
due to the uncertainty in orientation of the
dipole antenna, as discussed in Section 2.5.4.
If the angle between the axis of the projectile
and the normal to the ground (angle of in-
cidence) is a, then 0 may be any value in the
range (90-a) degrees to 90 degrees.

For most applications, the angle of incidence
a is less than 45 degrees. This means that the
terminal values of /2(0,<£) obtained with the
bar-type fuze, transverse excitation, are gen-
erally greater than those obtained with the
ring-type fuze. Furthermore, /2(0,<£) is gen-
erally a slower function of 6 in the region 45
to 90 degrees than in the region 0 to 45 de-
grees, so that the average values obtained
with the bar type depend less upon angle than
with the ring type. At any one angle of in-
cidence, however, the signal received by the bar-
type fuze on a particular projectile tends to have
a larger -spread than for the ring type, because
of the spread in dipole orientation mentioned
above.

Antenna Gain. The equation (93) shows
that M is proportional to the gain G of the fuze
antenna. The values of G obtainable with pres-
ent designs are relatively low, in the range 1.5
to 3. For an infinitesimal antenna G = 1.5, and
for a half-wave dipole G = 1.64. To get highly
directive antennas the frequency used would
have to be very much higher than used at
present, because of the small allowable physical
dimensions of the antennas. Because of this, the
antenna gain to date has not been a major de-
sign factor.

Carrier Frequency. From equation (103) it
is seen that the audio frequency of the im-
pedance change is proportional to the carrier
frequency. The amplitude of the impedance
change is also affected by the carrier frequency,
since M is proportional to A. Thus M is inversely
proportional to the carrier frequency, while the
audio frequency is directly proportional to the
carrier frequency. Besides these direct effects,
the carrier frequency affects M indirectly
through its influence upon the directivity pat-
tern (see Figure 23) and upon antenna gain. In
addition to these important effects upon the
impedance signal, it should be noted that the
carrier frequency also affects the circuit effi-

ciency and antenna matching, thus altering the
values of S that can be achieved. This, however,
is the subject of another chapter.

There is a discreteness in the possible firing
positions introduced by the use of the thyratron
control circuit. The spacing of these discrete
positions is determined by the carrier fre-
quency. Figure 29 shows how the discreteness
arises.

Figure 29. Illustrating the discrete character
of possible firing positions. Signal voltage is
plotted against height.

The solid curve represents the voltage out of
the amplifier which is gMS having the dotted
envelope gM0S0 . The horizontal line B represents
the holding bias. The point P, at which the en-
velope intersects the holding bias line, repre-
sents the point of function predicted by equa-
tion ( 105) . The fuze actually functions when the
M wave intersects the bias line at A a fraction
of a cycle later. The positive peaks of the M
wave occur for each A/ 2 reduction of the height.
For a particular fuze the size of the M wave
might be such that it just passes over one
positive peak. The predicted height of function
would lie near this peak but the actual function
would not occur until near the next peak about
A/2 further along. If the random variation of
fuze sensitivities is larger than the change from
one wave to the next, we will expect a random
height of burst with bunches located at posi-
tions roughly 1/2 apart. Thus the carrier fre-
quency determines the separation of the discrete
function positions.

This discreteness has been observed in field
trials. It implies that the average height of
function should be about (A/4) less than the pre-
dicted height.

WORKING SIGNALS; GROUND-APPROACH CASE

Ballistic and Target Factors

Distance from Target . It has already been
shown that the magnitude of M is inversely
proportional to the distance h of the fuze an-
tenna from ground. (This has been derived
from the consideration of the radiation field
alone and is not valid for distances close enough
so that the other components of the field are im-
portant. The modifying effect of these com-
ponents is considered in Section 2.10.) Thus a
plot of the resistance component of the reflected
impedance versus h will take the form of a
hyperbola, upon which is superposed a sinus-
oidal variation of space wavelength 1/2 and
another variation according to $he changes in
f2(6,<t>) as the fuze moves along its trajectory.
This latter factor is essentially constant over
the working range of any particular trajectory
for the ground-approach application.

Orientation of Fuze Antenna Relative to
Target. The part played by the directivity pat-
ten has already been described. The ballistics of
the situation, however, determine 6 and <f> and,
therefore, the use to which the directivity pat-
tern is put.

Because of a certain amount of rotation of the
projectiles in flight, <f> is generally indetermi-
nate ; for the ring type this is of no consequence,
since the directivity patterns have cylindrical
symmetry. The value of f2(0,<j>) for the bar-
type fuze suffers a certain amount of indeter-
minacy as we have seen.

For bombs released from flight, the angle of
incidence a increases as the speed of release in-
creases and as the height of release decreases.
This applies to dive bombing as well as release
from level flight.

For projectiles fired from ground, a increases
as the angle of elevation decreases. The speed of
projection also plays a part.

In all cases the ballistic properties of the
projectiles influence a in some degree because
of the effect of air resistance. Tables of inci-
dence angle a and vertical component of striking
velocity are too lengthy to be included here.
They may be found in standard bombing tables
or in special tables prepared for use with VT
fuzes.18'21* 70> 71j 74

Speed of Approach to Target. It has already
been shown that the audio frequency is pro-

portional to dh/dt, the speed of approach to the
target surface. This speed of approach is deter-
mined by the release conditions of the pro-
jectiles (angle, speed, and height) and by their
ballistic properties, and is essentially constant
over the last few hundred feet of flight.

Now it happens that a and dh/dt are some-
times so related that proper shaping of the
amplifier can be utilized to make up for wide
variations in a over a range of conditions, with
the result that the height of function remains
fairly constant over the range. This is con-
sidered in detail in Section 3.2.

Because of the usually high speed of ap-
proach to the target, the amplitude of the M
wave increases rapidly as the fuze nears the
point of burst. Since the impedance changes
are converted to voltage changes and then im-
pressed upon an audio-frequency amplifier, it
becomes important to take the dynamic char-
acter of the signal into account in order to
determine the output of the amplifier. This
matter has been the subject of considerable
study and will be discussed in Chapter 3. As
already indicated, it can be stated that to a
sufficiently good approximation we can usually
utilize the steady-state characteristics of the
amplifier in computations of the function point
of the fuze.

The terminal values of dh/dt encountered
in the ground-approach applications range
from about 200 to 1,200 fps.18 Over the range
of carrier frequencies 40 to 150 me, this repre-
sents a range in audio frequencies of 16 to
360 c. Only part of this range is encountered in
any one application.

Reflecting Properties of Target. The reflec-
tion coefficient n has already been defined in
Section 2.5.1. Using this definition the imped-
ance signal received from various types of
surface is proportional to n. Before going fur-
ther the method of measuring n for various
types of ground will be described briefly. The
apparatus consists chiefly of an antenna similar
to a longitudinally excited bomb, fed by a load-
sensitive oscillator. When the height of the
antenna above the reflecting ground is varied
the radiation impedance of the antenna changes,
these changes affecting the grid voltage of the
oscillator in a manner given by equations (89)

SECRET

THE RADIATION INTERACTION SYSTEM

and (93). The amplitude of the fluctuations
about the center value of grid voltage are re-
corded. These fluctuations are then compared
with those obtained in a similar experiment
with a large metallic screen for ground. Since
the metallic screen is practically a perfect re-
flector (n = 1), the effective reflection coeffi-
cient of the ground is given by the ratio of the
voltage fluctuations for the two types of reflec-
tor at the same height above each. Caution must
be observed, in making these measurements, to
use a large enough screen. The screen dimen-
sions must be large compared to the maximum
height used if complicated diffraction effects
are to be avoided.

The results thus obtained check well when
used to compute the actual magnitude of the
signal instead of ratios. They also check with
published values of the reflection coefficient for
plane waves.17’ 27

Table 3 shows the effective reflection coeffi-
cients of several types of surface.

Table 3. Reflection coefficient n.

Surface

Fresh water

0.8

Salt water

0.95

Average earth

0.5-0.6

Ice

0.2

If the ground surface is smooth, the value of
M is proportional to the reflection coefficient n
defined above. For the reflection coefficient n to
apply, the surface must be fairly homogeneous.
Irregularities, such as stones, that are small
compared with the wavelength will have little
effect upon M when the fuze antenna is at
least several wavelengths away. Areas, such as
puddles, that have a different reflection co-
efficient from the major part of the ground will
likewise have little effect, if they are small com-
pared to the height of function ; a sort of aver-
age reflection coefficient is involved in such
cases.

The effect of superposed targets, such as ice
over water, must be considered. Penetration of
the radio waves at the frequencies used is quite
small for metal, sea water, fresh water, dry
sand, and ordinary soil. Penetration into ice
or snow is considerably greater. Water a few
inches deep over a considerable area of land or

ice acts like a water target because of its small
penetration of the waves into water. A layer
of ice or snow that is only a few inches thick
gives a reflection coefficient more nearly that of
the surface beneath it than of ice or snow.

The effect of a general slope in the target
area is equivalent to a different angle of fall
over level ground.

The effects of large surface irregularities are
complicated and must be evaluated empirically.
When the fuze passes close to a large body, such
as a building, the reaction is similar to that
from an airborne target (treated in Section
2.11), and the burst occurs near the obstruc-
tion.

The effect of built-up areas like cities on the
height of burst is not yet well known. How-
ever, some general remarks can be made. The
general average reflection coefficient n will be
lower than for moist earth, i.e., about 0.4. In
general it is expected that the average height of
burst will be greater than that predicted on the
basis of the average n; the difference is about
half the average height of the structures. The
dispersion in height of burst will, of course, be
considerably increased.

The height of burst over densely wooded
areas has been found by experience to be just
below the level of the treetops for longitudinal
fuzes. Not much is known about transverse
fuzes under this condition.

When the fuze passes near the edge of a cliff,
it functions in a manner similar to the air-
borne-target case. When it passes over a bound-
ary between two different reflecting media such
as water and sand, there is a change in M
which is rapid if the fuze passes close to the
boundary and slower if the distance is larger.
Whether or not the transition causes a burst
will depend upon the transient response of the
amplifier and the abruptness of the transition
between reflectors.

2 10 EFFECT OF INDUCTION FIELD
ON CLOSE FUNCTIONS1

The preceding analysis of the reflected im-
pedance, based solely upon the radiation fields

f Bibliographical references pertinent to this section
are 12, 23-26, 39, 40, 94.

SECRET

EFFECT OF INDUCTION FIELD ON CLOSE FUNCTIONS

from the antennas involved, would indicate
that there is no change in reflected impedance
for the case of a longitudinally excited fuze
approaching the ground in a vertical direction
since f2(0) = 0. However, a little thought will
show that this is contrary to the principle of
conservation of energy. Therefore we must call
upon those fields in the vicinity of the antenna
which die away as the square of the distance
and the cube of the distance in order to de-
scribe the behavior of the fuze near the ground.

This can be seen as follows. If we use the
same argument as used in Section 2.14 with the
dipole oriented with its axis vertical, we find
that the total power radiated through the upper
infinite hemisphere varies with the height of
the dipole. These variations must appear as a
variation of radiation resistance and hence give
rise to an M signal. The higher-order terms
appearing in equation (160) (see Section 2.14)
represent the effect of these nearby fields for
the special case of the short horizontal dipole.
It should be stressed that the calculation is
made in terms of radiation fields alone but that
the results are identical with a treatment based
upon the induction and quasi-static fields.23

If attempts are made to extend this method
to the more complicated radiation patterns of
typical fuzes approaching at different angles,
the necessary integrals become impossibly com-
plicated and it is more convenient to use the
actual fields to calculate the result.

210 1 Second Approximation to the
Field Equations

In the previous discussion it was possible, by
restricting attention to 1/r radiation fields
alone, to avoid any reference to the current
distribution of the antenna and its coupling to
the feed system. In the argument which follows
it will be necessary to assume a current distri-
bution of a form which will give rise to the
observed radiation pattern and to consider the
interaction of this current with the reflected
fields. We shall be interested in the case of a
fuze approaching an infinitely reflecting ground.
The case of the airborne target does not, in
most cases, involve the nearby fields to a seri-

ous extent and has not been considered from
this point of view. As has already been shown
we deal only with the fuze and its image to
derive the necessary impedance change.

The problem thus becomes one of calculating
the mutual impedance of two identical antennas

Figure 30. Representation of fuze antenna and
its image.

oriented as shown in Figure 30. In this figure
we assume that the current distribution on
each antenna is given by

fi(^i) = hogM), (105)

-Ufe) = hogzfa), (106)

where I10 and /20 are the currents at the feed
points. We are forced to assume that the pres-
ence of either antenna does not alter the current
distribution on the other and that both are
identical, except for a 180-degree phase shift,
and the same as the distribution when each
antenna is radiating into free space.

Now the current /i(2i) gives rise to a field
#21(22) parallel to antenna No. 2 and vice
versa. Following Carter94 we write for the
mutual impedance

Z\2 = — j— f #21(22)02(22)^2. (107)

1 10 J

We first examine E2 1. For an infinitesimal dipole
of length dz, the field at distance r is given by

ES = bj^Sme[l-fr- - «,

(108)

E, = ^ cos 6 [- | - ^5] jet* - «,

#0 = 0,

where (3 = 2n/'k and b is a constant. At suffi-

jiSEC]

THE RADIATION INTERACTION SYSTEM

ciently large distances the higher orders of
(1 /r) can be neglected, thereby making Er
negligible and

Ee = -r— sin djei(fat ~ pr). (109)

It was this field that was used in the previous
analysis and called the radiation field, since it
accounts for the average radiated energy. The
other terms, in the integration of the Poynting
vector, give fluctuating components of energy
with no net energy flow.

The above expressions pertain to an infini-
tesimal antenna. The fields due to a finite an-
tenna may be obtained by regarding the antenna
as composed of infinitesimal antennas and inte-
grating the fields due to the individual infini-
tesimal antennas.

For a finite linear antenna placed along the
z axis, we have

(HO)

with a similar integration for Er. The integral
sign here denotes the limit of a vector summa-
tion over the whole antenna.

Now r and 6 are in general functions of z.
For points sufficiently distant from the antenna,
the dependence of d upon z is sufficiently slow
that it may be neglected. For comparatively
large distances the r dependence upon z may be
neglected in so far as the amplitudes of the
contributions of the individual elementary an-
tennas are concerned ; because of the finite
propagation time, however, the individual con-
tributions vary in phase. These phase variations
cannot be neglected. Taking into account the
above considerations, the components of the
electric field may be expressed as follows :

j(ut — 0 r)

Er = b

jpjiut — 0r)

[/ I Sr (/3r)2] '

sin d J'e-tt2™*0 I(z)dz, (111)

T- U - -1 1

(0r)\|

\-2J

cos d J'e-m*"** I(z)dz. (112)

In the above expressions the factor e -jpz cos e
takes into account the phase variations just dis-
cussed, since z cos 6 is the path difference for
contributions from z = 0 and z — z. The r in
the above equations is the distance from the
point z — 0 to the point P, where the field is cal-
culated. The term I(z) represents the z de-
pendence of the antenna current over its length
L.

If we denote the remote radiation field of the
finite antenna by ETad, equations (111) and
(112) may be modified as follows:

Ee [* ir (/Sr)2] Er°d’
Er = cot 6 E„

Era d je^wt ~ sin 6 j e ~^z cos 0

(113)

(114)

I(z)dz.

(115)

For thin antennas, sinusoidal current dis-
tributions are often assumed as engineering
approximations ; the integration may be effected
under this assumption, giving well-known
formulas for £rra(i for such cases.

For the fuze antennas, the current distribu-
tions are ordinarily not known with sufficient
precision to allow the carrying out of this inte-
gration. The 6 dependence of \Erad\ is obtained
experimentally by the method described in Sec-
tion 2.8.2. Thus the f(0) used there is given by

fid) = | sin 6 J' e~jPzcos0 1(z)dz |. (116)

The experimental method gives only the ab-
solute value as indicated and not the phase
dependence on 6 which is ordinarily not needed.

To get E21 from Er and Ee we use

En = Eoi sin 02 “f- Eri cos d2- (117)

The restrictions which allowed us to write equa-
tions (113) and (114) also imply that <92 = Oi
and that both are sufficiently independent of
z 1 and z2. Thus we write

E2 1 = Ee sin 6 + Er cos d = E21 e^1 ~ *>, (117a)

which defines E2i as a function which is inde-
pendent of z. In the exponential term r is the

EFFECT OF INDUCTION FIELD ON CLOSE FUNCTIONS

distance from z1 = 0 to the point 22 on antenna
No. 2. Thus the mutual impedance becomes

Z21 = — j— E21 J' e~JPr gziz^dzz. (US)
0

Again we must take account of phases, and so
we denote r as the distance between the point
zx = 0 on antenna No. 1 and the point z2 = 0 on
antenna No. 2. In this notation r = T z2 cos 0,
and we can further simplify equation (118) to
get

Z2i = —~E2ie~jp'r f e~^Z2COse g2(z^)dz2 . (119)

iio J

If we neglected all induction fields in the cal-
culation of equation (119), the only change
would be that E21' would reduce to E'raa sin 0
where F7'rad is defined in the same manner as
E2i. If we call the result of such a calculation
Z2 irad, we have

E*’

E' sin Q ’

rad

(120)

or, since M0 is proportional to Z21, we may write

M0 _ \E21'\

I E' sin ^ |

1 rad 1

(120a)

Comparison of equations (113), (114), (120),
and (121) shows that

V f1 ~ w? - m* cot29] + \jr + i cot2fl] ’

which reduces to

= T1 _ w2(4cot40_ x)]’ (121)

rad L- -1

after expanding and dropping inverse fourth-
power terms in r.

2102 Effect of Induction Field on
Function Heights

Equation (121) is a second approximation to
the calculation of Z12. It is not exact and is
limited in its application to cases where the

antenna is close enough to its image to make
the nearby fields appreciable but not so close
that the approximations used in deriving it
break down. Thus it can be expected to give
reasonably valid answers only for heights such
that the antenna length is not a considerable
fraction of the distance between it and its
image. Moreover, the whole derivation is based
upon a thin wire antenna as a model. The fat
antennas actually used may alter conditions.

Actually, in spite of the limitations, the
theory has been of considerable use in proper

Figure 31. Contribution of induction and quasi-
static fields to height of function; Ht as func-
tion of Hr for various values of 9.

selection of fuze antennas and the selection of
proper operating frequencies.

Since \dV\ is proportional to M0, we can then
say

dVr

(122)

where \dVr\ is the peak value of the voltage
change into the amplifier that would be com-

SECR

THE RADIATION INTERACTION SYSTEM

puted from radiation fields alone and \dVt\ is
the total voltage change including the effect of
nearby fields. If we now set \dVr | and \dVt\
each equal to the signal magnitude required to
actuate the fuze, B/g, the heights of function
predicted with and without the correction
terms, denoted by hf and hr respectively, are
seen to be related as follows :

s + ~ "■ (123)

This relation is shown graphically in Figure 31
for several values of 0. This graph shows the
contribution made by the induction and quasi-
static terms to the heights of function. This
contribution is seen to be significant when 0 is
small and when ht/k is small. For large values
of 6 and (ht/k) the correction becomes negli-
gible.

If we use Hr as the height of function, ex-
pressed in units of k as calculated without the
correction, and Ht as the height of function in
the same units and including the corrections, we
may write

h, = Hr[y2 + y2yJi + ' (124)

where

D =

4 cot4 0—1
4tt2

(125)

For a vertical approach Hr = 0, but the cor-
rection term becomes infinite leading to an inde-
terminate answer for Ht which must be evalu-
ated by further means.

From equation (104) we see that

Hr = ttnSoBf2(-e)- (126)

Also from equation (116) we see that

f\S) = | [sine / 1(z)dz |]2. (127)

Since the integral is a slowly varying function
of 6 for small 6 , we may write

P(6) = a sin2 6, (128)

for very small angles. For any particular pro-
jectile the value of the constant a is determined
experimentally from its measured directivity

pattern. Combining equations (124), (125),
(126), and (128), we get for very small angles

(129)

The latter value is independent of 6 and suffi-
ciently accurate for 0 from 0 to 10 degrees. It
is not valid for Ht < 1 ; in such cases the theory
does not hold anyway, as has already been
mentioned.

The equation (129a) is of interesting quali-
tative value, since it shows that the height of
burst at very steep approach angles depends
upon the slope a of the directivity pattern near
0 = 0 (slope on sin2 0 paper). Thus directivity
patterns like curves 5 or 6 in Figure 25 will
give very low height of burst even with the aid
of the induction field.

Figure 31 may be used to correct the heights
of function computed on the basis of radiation
alone. Referring to the example worked out in
Section 2.9, we have Hr = 20/8.2 = 2.4. From
Figure 31 we see that for 0 = 18.5 degrees and
Hr = 2.4, Ht = 2.7, or ht = 22 ft.

In this case the correction is not large. For
steeper angles of approach the correction be-
comes more pronounced. If radiation calcula-
tions predict a height of 2k for a striking angle
of 10 degrees from the vertical, the actual burst
height will be 3.5 k, a correction of +75 per cent.

A large amount of computation of heights of
function has proved to be necessary. For this
reason a method has been developed which
greatly reduces the amount of labor involved,
especially when rapid computations are needed
in order to help design an amplifier with a
shaping such as to give desired heights of func-
tion over a large range of ballistic and fuze
conditions. This method involves the use of
transparent charts.39

We see from equation (129a) that for steep
approach the height of burst varies as the
square root of the burst control factors, G, n,
S, g, and 1/B, and is thus less dependent upon
variations in these quantities when burst

SECRET

WORKING SIGNALS; AIRBORNE TARGET

heights are such that the induction field is
predominant.

A word of caution is necessary about incor-
porating amplifier delay to avoid noise troubles
on fuzes where the induction field is the pri-
mary signal controlling the height of burst.
Suppose, for example, we have a fuze on a
small antenna so adjusted that S = 15, g — 50,
and B = 4.5. Then for average ground n — 1/2,
and for the small antenna a = 1 and G = 1.5.
When these values are inserted in equation
(129a) the result is

Ht = 1.3.

an understanding of the operation. In Section
2.11.1 we shall consider the target to be a small
sphere which reflects equally in all directions.
First we consider the changes in phase in the
M wave and later the amplitude changes. Sec-
tions 2.11.2 and 2.11.3 will be devoted to a dis-
cussion of actual airplane targets. In all discus-
sions we use a reference system at rest on the
target.

2 11 1 Properties of the M Wave;

Simple Theory

If the firing level of the amplifier lags as
much as 3 c behind the calculated input firing
level, the fuze will reach Ht = Q before the burst
is initiated, since there are 2 c of input voltage
per 1 change in height. Such delay may give rise
to duds because the fuze breaks up before the
firing pulse is received. This behavior has been
observed in special fuzes carrying integrating
circuits to increase the resistance to noise pulses
and sweep jammer signals.

211 WORKING SIGNALS;

AIRBORNE TARGET^

The preceding two sections have been de-
voted to an analysis of the signals encountered
in the ground-approach case. The second im-
portant application of a radio proximity fuze
is to initiate bursts in the vicinity of an air-
borne target. In the present section the signals
occurring in this case will be studied.

To avoid complication we assume that the
fuze and target are far from ground and under-
stand that corrections may be introduced if
they are near the ground (see Section 2.5).

It is, of course, to be expected that the re-
flection from so complicated a target as an air-
plane will indeed be complex and not amenable
to exact analytic treatment. There are, how-
ever, certain general features of the problem,
which can be discussed in terms of a simple
model, that carry over into the actual problem.
A knowledge of these features is essential to

s Bibliographical references pertinent to this section
are 1, 14, 15, 55, 61, 72, 73, 75, 93.

Phase Properties

Figure 32 represents the spatial arrangement
of fuze and target (a small sphere). In this

TRAJECTORY

Figure 32. Spatial arrangement of fuze and
small spherical target.

figure the axis of the projectile is shown as
coinciding with the trajectory. This is, of
course, not time in the general case but is close
enough for the type of approach used in firing
rockets in air-to-air combat, since deflection
firing is seldom considered. The projectile is at
P and is moving along the x axis in the direc-
tion of the arrow. The distance between projec-
tile and target is r, and the shortest distance
between the target and the line of flight of the
projectile is p. The term p may be called the
impact parameter as in similar situations in
atomic physics. The line OQ is perpendicular
to PQ. The distance from P to Q is x. The angle
between PO and PQ is a. For fuze systems in
which the body of the projectile is used as the
antenna (longitudinal excitation), the angle a
is the same as the 6 previously defined. The
relative velocity of the projectile along its line
of flight is denoted by v.

SECRET

THE RADIATION INTERACTION SYSTEM

To proceed further, we recall that for the
case of reflection by a simple airborne target,
the M wave is given by a function of the form

M = M0e^~4^ + 5k (130)

There may be a phase shift at reflection which
will be a constant for the spherical target and
will be neglected.

The frequency F of the M wave is

F = ~ cos a. (132)

Thus the maximum possible value of the in-
stantaneous frequency is 2v/l when a = 0 de-
grees when the target is very far away, and the
minimum value is 0 when a = 90 degrees. As
the projectile approaches the target the fre-
quency decreases; when the projectile is at the
point of nearest approach to the target, the in-
stantaneous frequency is zero.

The rate of change of instantaneous fre-
quency with angle is given by

J_ I d/^irr\
2tt j dty X J

2 I dr I
X \dt\ 7

(131)

dF 2v .

— = — v si
da X

(133)

This equation indicates that the rate of
change of instantaneous frequency increases
as a approaches 90 degrees. For sufficiently
large values of a, the rate of change of fre-
quency with angle is great enough to cause a
considerable change of instantaneous frequency
during the course of 1 c. Thus it is important
to introduce dynamic considerations when
studying the response of an audio amplifier to
such a signal.

It should be pointed out that there are in gen-
eral only a relatively small number of waves of
M in that part of the trajectory where the fuze
is sufficiently near the target to be effective.
Calculation gives the result that there are
0.8 p/X waves in the region from a = 45 degrees
to a = 90 degrees.

For a typical case X = 8 ft, p = 50 ft, and
there are 5 c in the M wave. A longer wave-
length means fewer cycles and requires more
care in including dynamic considerations in the

study of the behavior of an audio-frequency
control circuit.

Caution must also be observed, in incorporat-
ing delay to avoid noise troubles and interfer-
ence from jamming signals, to make sure that
enough cycles are available to actuate the con-
trol circuit. Thus the higher the carrier fre-
quency the larger the number of waves avail-
able and hence the better a control mechanism
based on audio-frequency selectivity can be ex-
pected to function.

The manner in which the resistive component
of the reflected impedance changes as the fuze

THEOf

1ETICAL SIM

PLE ENVEL0I

/ / ‘

EDT

4 1 0

lLLLlX

OBSERVE

:0 ENVELOPE

0BSE

RVE0 M WAVE'' N

1 1

14 12 1

1 L

i i i i i i i i i

85* 40* 45* 50* 55* 60 * 65* 70* 75* 80*

Figure 33. Typical experimentally observed M
wave, with its envelope and theoretical simple
envelope.

antenna moves along its trajectory is shown
qualitatively by the dotted line in Figure 32.

Amplitude Properties

From equation (94) we see that

Af° = A -L Gf\a) cos r. (134)

Figure 33 shows a plot of M0 for a spherical
target with /2(a) the actual measured directiv-
ity of a typical fuze. It is marked “theoretical
simple envelope” on this figure. The ordinates
represent the M signal in arbitrary units. There
are two sets of abscissas, one set representing
—x/l, as defined in Figure 32, and the other set
representing a.

It will be observed from equation (134) that,
as the target is approached, M0 increases as the
square of the distance decreases and as the

SECRET

WORKING SIGNALS; AIRBORNE TARGET

directivity in the direction of the target in-
creases. We shall expect this general sort of
behavior even for a complicated target.

2,11,2 Reflecting Properties of Aircraft

When considering a complicated reflecting
target like an airplane we shall expect a varia-
ble phase shift upon reflection. This we repre-
sent by assuming that A, as defined in Section
2.5.2, is complex and of the form

A = F(r,d,<t>)eMr>0’*K (135)

At very large distances, such that the wave in-
cident on the target can be considered plane,
equation (135) will reduce to

A = (136)

the 1/r dependence of reflected field being taken
care of in the definition of A. At close distances
the whole argument becomes too complicated to
be within the scope of this report.

The variation in the phase of M arising from
5 (0,$) will change the spacing of the zeros
of M and thus alter the apparent instantaneous
frequency in a complicated manner.

It is helpful to consider the airplane as a
complicated antenna which is excited by the
incident radiation and which reradiates with a
many-lobed pattern characteristic of such an
antenna. Whenever the direction of the incident
radiation changes, the distribution of current
on the aircraft changes and the radiation pat-
tern is correspondingly altered. Moreover, if
the source of radiation is close so that the field
is not uniform over the target, the distribu-
tion of current will change with distance, also
giving a change in the reradiation pattern.

We might expect that the dependence of A
and 5 upon r will disappear for distances r such
that the target does not fill more than the first
Fresnel zone. We define the first Fresnel zone
in this case as that circular area such that the
path from the fuze to the center of the area is
X/4 shorter than the path from the fuze to
the outer rim. Experiments to be described
later show that this is roughly borne out.

For frequencies of 100 me and a target 50 ft
across, the target just bridges the first zone at

r — 125 ft. The radius of action of present-day
fuzes is well inside this range, so that simple
calculations can only give an order-of-magni-
tude effect. The actual values of M must be de-
termined by experiment.

There is, however, one very important factor
that minimizes the effect of the complicated
reflection on fuze performance. The factor
f2(a)/r2 grows so rapidly with increasing a
that there is a relatively small region in space
around the target where the signal is large
enough to actuate the fuze and the point of
burst becomes relatively independent of the
details of the wave form provided there are
cycles enough to get through the amplifier.

Experimental Measurement of Reflection
from Aircraft

The straightforward method of measuring
the reflecting power from an airplane is to con-
struct fuzes carrying radio reporting circuits
and fire them past an airborne airplane. This
gives the actual working signal but technical
difficulties make such tests impossible at the
present time.

For certain special interaction conditions the
airplane can be flown past the fuze, which is
held fixed in space. This gives information
which is useful for head-on or tail-on shots in
air-to-air combat; these are the most common
shots where rockets are concerned.

Experiments have been made for these con-
ditions in two ways. (1) The fuze was sus-
pended beneath a blimp and the airplane flown
by it. These tests gave reliable qualitative in-
formation about the signal voltages, but it was
difficult to get the exact distance measurements
required. (2) The fuze was supported at a
height of X/4 over a reflecting screen on the
ground and an airplane flown over it.1 These
latter experiments were made at the Naval
Proving Ground, Dahlgren, Virginia.

Properties of the Experimental M Waves

Some 50 space patterns were obtained from
which quantitative measurements could be
made. Parts of three typical patterns are shown
in the oscillograms of Figure 34. In Figure 33
a typical experimentally obtained pattern is
shown, together with its envelope, which is

THE RADIATION INTERACTION SYSTEM

compared with the theoretical simple envelope
discussed above.

An analysis of the phase properties of the
experimentally obtained waves gives very good
agreement with the simple theory given in
Section 2.11.1, showing that 5 (<9,<£) is a slowly
varying function and does not complicate the
pattern.

The experiments also showed that the inverse
square law holds. The upper curve in Figure 35

Figure 34. Oscillograms of parts of typical M
waves in fly-over tests.

These oscillograms were obtained in tests in which an
airplane flew over a fuze antenna mounted horizontally
above the ground. In the figures time increases from the
left to right. The vertical lines are timing pulses. The three
oscillograms are parts of a large photograph discussed on
page 23 of reference 1. Details of test conditions will also
be found in this reference.

shows the results of a series of tests in which
an SBD-1 airplane flew vertically over a fuze
antenna; the plane was in horizontal flight
parallel to the axis of the fuze antenna. The
fuze antenna was a distance X/4 above the
screenwire ground. The ordinate in Figure 35
is the logarithm of the magnitude of the signal
voltage AF, and the abscissa is the logarithm
of the distance p. It is clear from the figure that
the inverse square law holds for distances as
close as 80 ft for X = 7.4 ft as used.

From the Dahlgren experiments some quanti-
tative comparisons were made of the reflection
from the best aspect of the airplane and the
reflection from a tuned half-wave dipole. The
dipole was placed horizontally at various heights
above the fuze antenna and turned about a
vertical axis. As the dipole was rotated about
a vertical axis, the reflected impedance M0
changed from zero to a maximum value as cos r
varied from 0 to 1. The signal magnitude at
various heights was taken. Typical results are
shown in the lower curve in Figure 35. The
ordinate distance between the upper and lower
curves in Figure 35 represent a ratio of 21 ;

that is, the reflecting power of the airplane in
the aspect considered is 21 times as great as
that of a tuned half-wave dipole arranged for
maximum reflection. It has already been men-
tioned in Section 2.5 that the reflecting power
Amax of a flat plate of area L 2 is L2X, whereas
that of a tuned half-wave dipole is 0.26A. The
projected area of the airplane, regarded as a
flat plate, is about 300 sq ft. The wavelength X
used in this experiment was 7.4 ft. Then the
ratio

becomes 20.5, in satisfactory agreement with
the experimental values.

Figure 35. Signals from airplane and from
tuned half-wave dipole.

Expressing the reflecting power of the air-
craft in units of “dipoles” in the direction of
maximum reflection has been convenient in en-
gineering the fuze design. In the above de-

SECRET

WORKING SIGNALS; AIRBORNE TARGET

scribed experiment, the airplane is equivalent
to 21 dipoles.

The envelope of the wavy curve in Figure 33
gives a rough idea of the variations in M0 as
the position of the fuze antenna changes with
respect to the airplane target. This wave repre-
sents the signal received by a radio fuze on a
rocket due to a small airplane passing the
rocket at a distance p equal to about 15A or
105 ft in this case. To indicate the extent to
which the signal acts as would be predicted
from a point target, an envelope computed ac-
cording to the simple theory is plotted on the
same graph. That is, the simple envelope is a
plot of the equation

M0 = Constant X - (137)

whereas the observed envelope is effectively a
plot of the equation

Mo = Constant X ^ . (138)

The two envelopes are adjusted so that their
maximum values coincide. The observed enve-
lope indicates the nature of the variation in A
along the trajectory. It is clear from the figure
that the simple features of the theoretical en-
velope are modified in the observed envelope,
in that there is a minimum around a = 60 de-
grees, whereas the amplitude of the theoretical
envelope rises constantly as a increases toward
90 degrees.

Other patterns experimentally obtained for a
variety of aspects and distances exhibit similar
features, that is, a general trend as expected
from the simple theory, plus a superposed effect
of one or two minima. The positions of the
minima vary with aspect and type of airplane.
Thus the amplitude properties of the wave fol-
low the simple theory to a certain extent ; fur-
thermore, as we have seen, the phase properties
follow the simple theory quite well. Thus the
characteristics of the wave are sufficiently well
known to achieve good burst control, as will be
shown in Chapter 3.

Specification of Sensitivity Requirements
for Plane-to-Plane Rocket Fuze

We are now in a position to outline a
sample calculation for a trial design center for

a fuze for the plane-to-plane rocket application.

The fuze will fire when

M. - A (139)

Within engineering limits, the quantity B/S0g
can be varied at will, so we shall calculate the
value of M0 to be expected and leave the discus-
sion of B/Sog to appropriate chapters. The
maximum value of M0 has been seen to be equiv-
alent to about 20 dipoles for the direction of
best reflection from the airplane. We take the
value of 10 dipoles as a working average.

From equation (49) we see that A for a sin-
gle dipole is 0.26A, and with the aid of equation
(134) we find the reflection from a target of
10 dipole strength to be

Mo( 90°) = ^p/2( 90°),

at the point of maximum reflection from the
target. For a typical fuze with a half-wave an-
tenna G = 1.64, /2 (90°) = 1, and X = 7.5 ft.
If we wish the fuze to function up to a distance
of p — 10X,

M0( 90°) = 2'62^-'-4- = 0.007.

If, however, it is desired to have the burst occur
when a is approximately 60 degrees, for an as-
sumed maximum fragmentation density in that
direction, /2(60°) is about */2 and

r2 = _eL. = lE2

sin2 a 3

These values make

M0{ 60°) = 0.0026. (140)

Experiments have shown that the loss asso-
ciated with the dynamic response of shaped
amplifiers is about 10 per cent. Thus we reach
the final conclusion that a workable fuze must
function when

Mo ^ 0.0025. (141)

To refine the calculation further would be
useless, since the answer is only approximate.
Trial fuzes were built to fire when M0 = 0.0025
and tested against a mockup target. It was

found that this value gave good field results,

the final proof of any design.

THE RADIATION INTERACTION SYSTEM

It will be noted that the reciprocal of the
value of M0 required to fire the fuze is a meas-
ure of the overall sensitivity of the device. It is
convenient, as an aid to thinking, to specify this
value in terms of a special type of test. We
imagine the fuze moved toward an infinite per-
fect reflector at such a speed that the doppler
frequency is exactly right for maximum ampli-
fier gain. We also assume the projectile to be so
oriented that the direction of maximum radia-
tion is toward the reflector. Under these special
conditions

M o =

XG
4 irk’

(142)

and the fuze will function at a height h, given by

ItWo

(143)

For the calculation just outlined above this re-
duces to

^eff —

7.5 X 1.64
4tt X 0.0025

^ 400 ft.

For a quick specification of the overall fuze
performance this effective height is quite con-
venient. This method of specifying the sensi-
tivity of a fuze is called the Michigan sensitivity
and was used extensively by Section T, Office of
Scientific Research and Development.

Experience resulting from tests against a
mockup target as well as against actual targets
has shown that fuzes with hett = 400 ft are
quite satisfactory, but that greater sensitivity
can be used with increased effectiveness. Values
as high as 800 to 1,000 ft have been achieved in
later models.

The effective height as defined above should
be used only for comparing the effectiveness of
fuzes working on a given frequency. As seen
from equation (143), hett is proportional to X.
On the other hand, for an airplane target, M0 is
not proportional to X but to a first approxima-
tion is proportional to XL Thus for smaller X,
hett is reduced while performance against an
airborne target is not reduced in proportion
(see Section 2.11.3 following).

21 13 Effect of X on Reflection from

Aircraft

Equation (134) shows that for all other con-
stants equal

M0 ~ AX. (144)

Mott93 has calculated the values of A for vari-
ous simple reflectors and finds

Dipole: A ~ X1.

Sphere: A ~ X°.

Flat plate: A ~ X-1.

Considering all factors, he recommends as a
working average value A ~ X~L Equation (144)
then becomes

M o ^ X*

as an average case and M0 ~ X° for the case of a
flat sheet like an airplane wing. Our fuze ex-
perience indicates that the behavior is more
nearly the latter than the former.

SIGNAL SIMULATION11

2.12.1 Properties Required of Simulator

In the preceding sections a description of the
impedance signal due to a reflector has been
given, and it has been shown that the changes
in amplitude and phase of the impedance sig-
nal M are duplicated as amplitude and phase
changes in dV.

The value of M is a function of position alone
and can be represented as a space pattern along
the trajectory. This space pattern has been
called the M wave. The value of M as a func-
tion of time is obtained once the position is
known as a function of time. The form of the
M wave is not altered by the velocity of ap-
proach; the wave is merely traversed at an
appropriate rate. This means that it is possible
to measure the M wave point by point with
static impedance measurements and compute
its time variations in any given case from the
specified relation between position and time.

h Bibliographical references pertinent to this section
are 3, 7, 35, 46, 76, 92.

SECRET

SIGNAL SIMULATION

If we are to simulate the working impedance
signal, we must devise an arrangement which
presents to the fuze circuit an impedance which
has the correct amplitude and time dependence.
We are not concerned here with a device which
attempts to simulate the vibration and stress
conditions encountered by the fuze in actual
use.

Experience has shown that the r-f part of
the fuze system is able to respond to changes
of antenna impedance much more rapidly than
any encountered by the fuze in practice. This
means that the audio-frequency circuit is the
part of the fuze that responds to velocity
changes. As a result of this division of func-
tion, it has been found desirable to test the
r-f system and audio system separately in en-
gineering the fuze design. A final test which
simultaneously measures the combined per-
formance of the complete system serves as an
overall check to insure that there are no unde-
sirable interactions.

A truly faithful simulator must actually re-
produce the rotating impedance vector which
is characteristic of the interaction with the
moving target. However, it will be seen in
Section 3.1.7 that the sensitivity of the r-f cir-
cuit to reactance changes is usually negligible in
comparison with the sensitivity to resistance
changes, so that a simulator which reproduces
the resistance component of M is adequate for
most work. This simplifies the simulator prob-
lem greatly but it must always be remembered
that an approximation is involved when such a
“resistance simulator” is used.

Additional properties that a simulator must
possess are those which make its use practical.
It must have a sufficient range of operating
conditions, it must be reproducible, it must be
convenient to use, and it must not introduce
complicating disturbances into the fuze circuit.

2 12 2 Field R-F Simulator

There is one convenient method of simulation
of the true antenna impedance variation that
involves the use of a field setup. It is in fact not
a simulator in the true sense of the word, since
it reproduces the actual voltages as a function

of distance from the target but not on the
proper time scale. It is, however, discussed here,
since it is one method for presenting the proper
antenna variations to the r-f system. We refer
to what is commonly called a “pole test.”

In making pole test measurements, a fuze is
mounted in a mockup of the proper projectile
and suspended over a large reflecting screen by
ropes attached to tall poles. The height above
the reflector is varied, and point by point read-
ings of the voltage ciV are recorded. The re-
quirement for proper antenna simulation is
automatically fulfilled.

If the readings are taken at a height of sev-
eral A, M0 is quite small, and we can say that

(145)

except for a fixed phase shift which has been
neglected. If calculated values of M0 are sub-
stituted in this expression, S0 can be calculated
for a given r-f circuit.

In practice it is difficult to obtain the read-
ings at a very large height so that M0 may be as
large as 6 to 10 per cent. It can be shown that
the error in S0 calculated in this manner is
about the same size as M0. Such measurements
are accurate enough for most purposes.

There is apt to be a large error in pole test
measurements if the reflecting screen is not
large enough. When the height of the fuze be-
comes comparable with the semidiameter of the
screen, diffraction effects set in which are of
unknown phase and magnitude. Errors as large
as 100 per cent have been observed. It is de-
sirable to have the screen diameter at least four
times the height of measurement.

2 12 3 Laboratory R-F Simulators

By laboratory r-f simulators we mean those
devices which generate impedance changes of
a form suitable for making tests of the com-
plete r-f system, but which do not have the
amplitude-time dependence necessary for test-
ing a complete fuze system. They fall into two
general categories, those which vary the re-
sistance component alone and those which set
up the true rotating impedance vector. The

dV = M0Sq sin I

: SECRET \

THE RADIATION INTERACTION SYSTEM

latter, while interesting, do not give enough
additional information on present-type fuze cir-
cuits to justify their construction. Both types
will be discussed briefly. Since the fuze circuits
used at the present time have small reactance
sensitivity, there has been no need for react-
ance simulators, and none has been designed.
However, as will be pointed out, the reflecting
dipole simulator can be used as a reactance
simulator if desired.

Resistance Component Simulators
(Substitution)

This is the simplest of all tests to make in the
laboratory. One merely disconnects the fuze
antenna from the circuit and adds a dummy
antenna consisting of lumped resistors and
condensers, which duplicate the operating point
of the r-f system. The resistance is varied by
substituting several resistors in succession. A
curve of V versus In Rp (or In Rs) is plotted, and
the slope of this curve at the operating point is
the sensitivity Sp (or Ss).

If desired, the fuze assembly can be placed
inside a shield can instead of removing the an-
tenna exciter. Such an arrangement leaves the
operating point reactance practically unaltered,
and resistors can be substituted directly across
the feed point. This latter method is preferable,
since any stray coupling between oscillator and
antenna is left undisturbed.

Figure 36 shows a typical load curve obtained
from such a series of measurements. It is a
curve of V versus Rp on a logarithmic scale.
From such a curve Sp, which has been defined
as dV/d In Rp, may be found.

Resistance Component Simulators
(Dipole Reflectors)

The dipole reflector is not strictly a resistance
simulator, since it can be made to perform as a
rotating vector simulator, resistance simulator,
reactance simulator, or combination simulator.

We may see with the aid of equations (48)
and (94) that the reflection from a half-wave
dipole oriented so that /32 (03i,4>3i) = 1, is given
by

M = cos’

(146)

We see that M can be changed by changing r,
Z33, /i2(0i3,</>i3) , and t. All these changes have
been utilized at one time or another. The dipole
is adjusted so that

Z33 = (Rsz + Zl), (147)

where ZL is the external impedance connected
to the feed terminals of the dipole. If we short
the terminals, ZL = 0, and Z33 reduces to Rs3.

Wand. A wand consists of a length of wire
cut to resonant length. In effect ZL = 0, since
the terminals are shorted. It can be used in two
ways. First, it can be oriented so that t — 0 and
moved toward or away from the fuze antenna,
thus varying r. When so used the signal pre-
sented to the fuze is truly the rotating vector,
and the changes dV can be recorded as the
wand is moved.

Second, it can be so located that M is purely
resistive, t may then be varied by twisting the
dipole in a plane perpendicular to r, thus vary-

I 2 3 4 567 89 10 20 30 40507090

R THOUSANDS OF OHMS
P

Figure 36. Typical loading curve.

ing the effective resistance. If desired, the
dipole can be attached to a motor so that t = co t.
Then an audio signal will be developed which
varies as cos2 co£.

SIGNAL SIMULATION

If reactance simulation is required, r may be
adjusted so that M is purely reactive and the
rotation gives a reactance variation instead of
a resistance variation.

Modulated Dipole . The term ZL can be varied
by connecting a variable impedance to the ter-
minals of the dipole. The variable impedance
can be provided by a rotating condenser, com-
mutator, flashing thyratron, or any other con-
venient form of variable r-f impedance. Unless
the variable impedance can be made purely re-
sistive or purely reactive, the resulting M be-
comes quite complicated because time-varying
phase shifts arising from changes in ZL are
included in the reflector.

When dipole simulators are used in the lab-
oratory, stray reflections set up complicated

Figure 37. Basic circuit for diode simulator.

standing waves in the room, and only qualita-
tive answers are obtained. A change of position
of the dipole with respect to the fuze varies
f2(0,<f> ) and gives crude information about the
directivity of the fuze.

Such devices are useful for quick checks to
see if fuze circuits are “live” and have approx-
imately the desired sensitivity. Because of the
(k/r)2 factor in the reflection the device can
not be used effectively at a large distance from
the fuze and hence cannot be used well for
measuring directivity patterns.

Resistance Component Simulators (Diode)

The a-c input resistance of a linear diode
voltmeter is a function of the d-c resistance and
d-c voltage in its load circuit. The fundamental

circuit for a diode simulator is shown in Fig-
ure 37.

Terminals TT are connected to the antenna
terminals of the fuze circuit. The reactance X
represents lumped reactance (including a d-c
return, if necessary) to make the input react-
ance of the device simulate the operating point
reactance of the fuze antenna.

By varying v and RL the apparent r-f re-
sistance R between terminals TT can be con-
trolled. In practice RL is adjusted to give the
value of R at the operating point when v = 0.
The term v is then varied at an audio-frequency
rate to introduce small periodic variations in R.

If the diode is working in its linear region so
that a d-c voltmeter indicates a voltage V across
Rl of several volts, the simple diode theory
works quite well. Let Rp be the dynamic re-
sistance of the diode and 6 the semiangle of
flow when v = 0. If v/V«l it can be shown
that

dR vR . v ,

y^^(2 cos 0) -y p, (148)

where 6 is determined by

Rp 6 — cos 6 sin O’ y '

and subsequently Rl is determined by

= - (tan 6 - 0). (150)

■tl L TT

Figure 38 shows the appropriate values of RL
to give the desired value of R when Rp is known,
and the value of Pf the correction factor that
must be applied to v/V to give dR/R.

It is not wise to use values of R/Rp less than
10. If lower values of R are needed, a fixed re-
sistance can be shunted across the circuit, with
X and the ratio of v/V adjusted accordingly to
make the overall dR/R have the desired value.

The diode simulator need not be connected
directly to the antenna terminals but may be
capacitatively coupled, if appropriate calcula-
tions are made and provided the diode is oper-
ated in its linear region.

In laboratory work it is advantageous to cali-
brate the simulator directly by attaching it to a
stable fuze circuit which has a known load
curve. The desired operating point is selected

THE RADIATION INTERACTION SYSTEM

and the output of the fuze circuit measured as a
function of v/V. From the measured output and
the known load curve the effective dR/R for the
simulator can be calculated directly with no
detailed knowledge of the diode. The theory
above serves as a useful guide in selecting
proper diodes and in indicating the range of use-
fulness of a given simulator.

The diode simulator has a considerable
advantage over thermistor-type simulators

because there is no delay in response to the
applied audio voltage. Tests have shown that
the device will follow frequencies far in excess
of any required in fuze testing.

Resistance Component Simulators
(Thermistors)

The effective resistance of an r-f circuit can
be controlled by incorporating a thermistor ele-
ment somewhere in the network. The temper-
ature of the thermistor can be varied at an
audio rate by passing audio-frequency current
through it. This results in a variation of the
effective resistance of the r-f circuit at an audio-
frequency rate.

Typical thermistors that have been used are
small flashlight bulbs and Littel fuzes. Because
of the thermal lag of such devices the upper
audio frequency is quite limited, and each de-
vice requires calibration against a standard
fuze circuit of known stable performance.

A detailed description of a thermistor simu-
lator and its calibration will be found in an
NDRC report,3 which shows the general pro-
cedure for calibration of any resistance simu-
lator.

Resistance Component Simulators (Triode)

The effective resistance of an r-f circuit can
be controlled within limits by loading it with
a triode so arranged that the dynamic plate
resistance of the triode is used as an r-f re-
sistance. The value of the dynamic plate resist-
ance can be changed by changing the grid to
cathode potential of the triode. The changes in
plate resistance respond to changes in grid
voltage at frequencies far greater than any
needed in fuze testing. Hence this device com-
pares favorably with the diode simulator. Cali-
bration is necessary, and in general the circuit
arrangement is more complicated than for an
equivalent diode simulator. The details of a
typical triode simulator developed by the Philco
Corporation are shown in their final report.70

Rotating Vector Simulators

There has been little need for true rotating
vector simulators aside from the pole test, which
serves as a final check on any fuze circuit. In
some special phases of fuze work, however, a
rotating vector simulator is of interest.

Several schemes have been proposed and two
put into practice. One of these is a side-band
type by Airborne Instruments Laboratory for
use in their countermeasure studies; another
has been designed by Westinghouse for testing
a pulse type of fuze.

The first type receives a carrier from the
fuze, adds two side bands at simulated doppler
frequency and cancels out the carrier plus the
lower side band. The upper side band is ampli-
fied and reradiated to form a true rotating
vector simulator when mixed in the fuze cir-
cuit. The details will be found in an NDRC
report.92

The second type uses a transmission line with
two resistance simulators located at points sep-
arated by (1/8). The audio drive on one simu-
lator is 90 degrees ahead of that on the other.
The resultant effect of the two is a rotating
impedance vector at the input to the line, when

SIGNAL SIMULATION

the load presented by each simulator is prop-
erly adjusted.

Laboratory Audio Simulator

As has already been pointed out the voltage
wave into the amplifier is of the form

dV = MS, (151)

and S is essentially a constant of the r-f system.

forms of M were sufficiently simple, laboratory
sine wave oscillators could be used for all audio
circuit tests.

The rate of change of instantaneous fre-
quency and amplitude of the M wave for an
airborne target are so large that it is extremely
tedious to predict amplifier performance on the
basis of its steady-state response to sine waves
of various frequencies or on the basis of its
transient response to a unit pulse.

It has been found very convenient to circum-

Figure 39. Audio-frequency M- wave simulator; view of drum and associated equipment.

The control of burst and discrimination against
noise are performed in the amplifier control
section of the fuze. In testing this part of the
system, it is not necessary to include the r-f
elements, provided an audio voltage wave pro-
portional to M can be generated. If actual wave

vent these difficulties by constructing an audio-
frequency simulator which generates a wave
which is in detail like the wave measured in the
fly-by tests described in Section 2.11.3.

In principle the device is quite simple. The
desired wave form is cut on an opaque paper

THE RADIATION INTERACTION SYSTEM

tape and wrapped around a transparent cylin-
der. A light source is placed inside the drum.
It illuminates a tiny transverse strip of the
tape by means of a slit. A photocell on the out-
side of the drum measures the light passed by
the tape. When the drum is rotated by a motor
drive, the M wave of voltage is generated.

Since the form of the M wave does not de-
pend upon the speed of the projectile, the same

Figure 40. Audio-frequency Af-wave simulator;
view of control panel.

wave can be used for all projectile velocities.
Thus the speed of rotation of the drum corre-
sponds to projectile velocity, and a whole range
of interaction velocities can be simulated with
a single adjustment.

If an oscilloscope sweep is synchronized with
the drum rotation, it becomes a simple matter
to investigate delay in circuit response. By in-
corporating several channels, noise of typical
forms can be superimposed on the M- wave sig-
nal to demonstrate the discriminating proper-
ties of audio control circuits.

Figures 39, 40, and 41 show photographs of

a typical audio simulator, and oscillograms ob-
tained with it.

The audio simulator is used also in studies of
the ground-approach M wave. Although this
wave form is not so complicated as that from
an airborne target, the rate of rise of amplitude

Figure 41. Simulated M wave obtained with
audio-frequency M-wave simulator, with super-
posed response of amplifier to simulated M wave.

In each photograph the curve of larger ampli-
tude represents the M wave. The superposed
amplifier response is scaled down. In the upper
photograph an amplifier peaking at 100 cps was
used ; in the lower photograph the amplifier
peak was 50 cps.

is large enough to make dynamic studies of the
amplifier necessary. These are performed more
easily on the audio simulator than by calcu-
lation.

The audio simulator has proved itself to be
a worth-while research tool and may be of con-
siderable use for other laboratory work.

ANTENNA NOISE

2.12.5 Overall Signal Simulator

In case it is necessary to simulate at the an-
tenna terminals the complete variation of im-
pedance, a combination of the audio simulator
with the resistance simulator of the diode or
triode can be used. The voltage from the audio
simulator is used to drive the r-f resistance sim-
ulator. The result is a wave which presents the
correct variation of radiation resistance to the
fuze circuit. Reactance changes will not be in-
cluded, but they are not normally needed.

2 13 ANTENNA NOISE

2 131 Introduction

We have seen in equation (44), reproduced
here for convenience,10* 42 45 that the presence of
a reflecting body changes the input impedance
of the fuze antenna, thus

Zx = Zn - Zr. (44)

We in effect altered this equation to read

and tacitly assumed that Zu is constant so that
the only changes in Zi are those produced by M.
We have also seen that expected values of M0 are
very small (~ 0.0025 for the airborne-target
case) and that fuzes are designed to work on
these small changes when they have a proper
time dependence.

Now a fractional change in Zn will be just
as effective in actuating the fuze as the whole
of M, if it has the proper time dependence. Such
a change is indistinguishable from the expected
signal, and the fuze will function if these
changes in Zu occur.

There are two physical differences between
the Af signal and a variation of Zn. These are
(1) the time delay associated with the time of
flight of the radiation to the target and back
and (2) the fact that the M signal is a return-
ing wave instead of an outgoing wave.

Up to the present time no practical schemes
have been evolved for making any differenti-
ation. Such schemes will no doubt be developed,

but the fuzes with which this report is con-
cerned cannot make the distinction. The con-
ventional radar pulse-time system makes the
necessary distinction but has not yet been de-
veloped in a form suitable for small fuzes.
Hence a variation of Zn gives rise to a signal
of the same form as the M signal.

Unfortunately circumstances arise wherein
Zn is not constant, and we are forced either
to rely upon the difference in time variation to
discriminate between M and variations in Zn
by means of the audio control system or to go
to severe lengths to hold ZX1 adequately con-
stant. There are two sources of variation in Zu.
These are (1) geometric deformations of the
antenna structure associated with vibration, and
(2) erratic additions to the antenna system
arising from propellant flames associated with
the projectile itself. The latter source of trouble
is associated only with self-propelled projec-
tiles such as rockets or guided missiles.

It may be mentioned that the problem of
thermal noise never arises, since the signal
levels used are always much higher than the
thermal noise level.

Signals originating in radiators other than
the fuze are considered interference, and the
susceptibility of fuzes to these signals is not
primarily an antenna problem but rather an
internal circuit problem. The antenna plays a
small part by virtue of its reception pattern,
effective length, and polarization. These prop-
erties have been discussed in preceding sections
and need not be considered further.

The whole interference problem is intimately
related to the problem of countermeasures for
the fuzes and is therefore not treated in detail
here. We now turn attention to the antenna
noise as defined above.

2.13.2 Antenna Noise Resulting from
Geometric Deformations

The normal dimensional deformations asso-
ciated with vibration in projectiles are so small
that they can be neglected. The real trouble
arises when the vibration varies the contact re-
sistance (or impedance) between parts of the
projectile. This may occur between parts of a

THE RADIATION INTERACTION SYSTEM

welded-fin structure on bombs, at the point
where the tins are attached either on bombs or
rockets, at the point where the fuze is attached
to the projectile, and at the point where the
power vane is attached to the fuze.

The obvious solution to the whole problem is
to make all joints so tight electrically that the
variations do not matter. Usually it is possible
to achieve the required tightness provided (a)
the fins are made properly, and (b) the assem-
bly is tight when the projectile is used. It is
difficult, if not impossible, to simulate in the
laboratory vibration conditions like those set

Figure 42. Two possible current distributions
on fuze antenna.

up in the actual projectile. Thus the only ex-
perimental method of determining when the
desired degree of tightness has been achieved
is to make field tests with real projectiles carry-
ing fuzes known to be internally quiet. Such
experiments are made with each type of pro-
jectile to be used, for “proving-in” purposes.

The antenna noise generated by the rotating
power vane cannot be removed by tightening
the system. It can be reduced by putting an
electric shield around the propeller, as in the
case of the ring-type antenna, or by using a
rotating system whose speed is so high that
the noise frequency is higher than the expected
doppler frequency. The audio control system
can then discriminate between the doppler fre-
quency signal and the antenna noise. This latter
device is used in both bar- and ring-type fuzes.
The erratic noise set up in the nose bearings is
reduced to an acceptable level by keeping the
amount of metal in the rotating system exposed
to r-f fields very small.

When tactical conditions and engineering de-
sign considerations permit, the noise level can
be reduced by an appropriate choice of carrier
frequency. To see this consider the schematic

antenna system shown in Figure 42. The dif-
ferent standing waves of current are shown by
the dotted lines. The single-lobe pattern corre-
sponds to a frequency such that the antenna
is about (A/2) long. The double-lobed pattern is
that for a frequency such that the antenna
is nearly l long. Suppose that the contact re-
sistance varies at point x. The current through
x is larger for the l wave than for the {1/2)
wave. Hence the power absorbed at x varies
more for a given variation in x when the l pat-
tern is used than when the (A/2) pattern is
used. Power absorption appears as variations in
Z? i and hence appears as a spurious signal.

If the noise point happened to be at y, the
l pattern would be better. Usually the fuze
joint is near one end and the fin joint near the
other, and it is not possible to get a current
pattern which places nodes at these points.
Furthermore, fat antennas do not have marked
nodes except at the ends. For this reason it is
generally better to use a low carrier frequency
to suppress vibration noise.

This argument has been verified in the field
in the case of longitudinally excited 500-lb
bombs.

2.13.3 Antenna Noise Resulting from
Propellant Flames

Rockets are particularly subject to this type
of noise, since they carry a long flame behind
them for a considerable portion of their flight.
It is possible to delay the arming of the fuze
until the propulsion blast is over without im-
pairing the effective use of present-day rockets
too greatly. However, the trend is toward longer
burning rockets and the flame is sure to become
a serious problem. Furthermore, there is more
to the problem than first appears. When the
burning is over, the flame does not go com-
pletely out and remain so. Scraps of unburned
propellant left in the hot motor reignite and
give small “chuffs” of flame which do not dis-
turb the motion of the rocket but which do
disturb the behavior of the antenna. These
chuffs of flame have been found to occur er-
ratically many seconds after the main burning

SECRET

ANTENNA NOISE

of the propellant has ceased. They create large
enough changes in ZX1 to cause the fuze to func-
tion before it reaches its intended target. The
problem of afterburning, as the phenomenon of
the chuffs has been called, has been a serious
one in the case of present-day rockets, and con-
siderable effort has been expended in seeking
a solution of the problem.

The attack on the problem has taken two
main lines. These are as follows:

1. A study of the electric properties of the
flames to see how circuits can be designed to
suppress the response. Such electric properties
are within the scope of this chapter.

2. A study of the means for eliminating the
afterburning problem by stopping the after-
burning. This phase of the attack is treated in
another chapter, since it is not an antenna
problem.

The electrical effects of the flame result in
the production of a spurious signal which can
be distinguished from the expected M signal
only by its different time variation. It has been
a simple matter to show that the flame does
actually produce large spurious signals. A
rocket carrying a fuze was mounted on an in-
sulating stand and connected through insulat-
ing hose to supplies of gas and air. By this
means flames of any desired size could be pro-
duced at the end of the projectile. Recording
instruments were connected to the fuze in such
a manner as to leave its radiating properties
essentially undisturbed.

Arrangements were also incorporated so that
the flames could be started or stopped quickly.
The sudden change gives rise to time-dependent
effects which can be easily separated from the
steady-state r-f conditions in the absence of
the flame. Several interesting properties of the
flames were immediately evident.

1. The yellow sooty flames from pure illumi-
nating gas had no measurable effect.

2. The clear blue flame from a mixture of
gas and air had no effect.

3. When arrangements were made to spray
NaCl, or KC1 solution or powdered salt into the
flame, a large response was observed immedi-
ately. The changes were five to ten times those
needed to trigger the fuze normally.

4. The effective flames were not in contact

with the fuze, being separated from it by an
inch or more of nonburning nonionized gas.

5. The magnitude of the effect could be
changed by changing the concentration of the
ionizing substance injected into the flame.

6. The magnitude of the effect could be
changed by changing the flame length.

Figure 43 shows a typical curve of signal
versus flame length. The ordinate is expressed
in terms of the value of M0 that would be re-
quired to give the same signal. The arrow near
the base at 0.0025 represents the working value
of M0 for which the fuze was designed to fire
when the proper signal is approached.

It was next necessary to demonstrate that
the rocket propellant actually carried enough

0 10 20 30

FLAME LENGTH (INCHES)

Figure 43. Signal produced by flame, as func-
tion of flame length.

ionization to d,o what the flames did. To check
this a rocket was supported on a strong insu-
lating support, and voltage changes measured
while the main burning was going on. Voltages
larger than in the above described experiment
but of the same order of magnitude were ob-
served. The effect of afterburning was checked
by putting small amounts of propellant near
the nozzle of the motor and igniting them with
a hot wire. The signals from the lower-temper-
ature burning were still of the same order of
magnitude.

These tests leave no room for doubt about the
signal-producing properties of a flame. Field

SECRE'

THE RADIATION INTERACTION SYSTEM

tests have shown that such flames do exist and
that they are associated with fuze functions.

Studies of circuit behavior were undertaken
to see if other carrier frequencies might be use-
ful. Reasonable changes which could be readily
incorporated into the fuze design were investi-

Figure 44. Diode voltage versus antenna length.

gated. None of these changes eliminated the
effect of the flames.

Figure 44 shows curves of the d-c voltage
output (diode voltage in this case), from the
r-f section as a function of the length of the
projectile-antenna. Time-dependent changes in
this voltage represent the dV set up by the
M wave. In effect the curve may be considered
to be a plot of Rv although it is not exactly pro-
portional to it.

To get the curves a piece of brass pipe was
used to simulate the rocket. The upper dotted
curve in this figure shows the voltage as a func-
tion of the length of brass pipe. The length of
33 V2 in. corresponds to the length of the rocket
for which the fuze was designed.

The lower two curves show the response when
the length in excess of 33!/£ in. consisted of a

paper tube coated with Aquadag. The coating
was found to have a d-c resistance of 20 ohms
per in. for the upper curve and about 100 ohms
per in. for the lower curve. These curves indicate
the effect of an extension to the antenna which
is not a very good conductor. That is, it simu-
lates more nearly the conditions of the conduct-
ing flame. We see that the lengthening of the
antenna gives voltages which are very large
compared with the 30-mv signal required to fire
a fuze. They are larger than those observed
from the flames. This was thought to be due
partly to the air gap between the flame and the
antenna.

To show that the size of the gap makes a
considerable difference in the effect of the ex-
tension, two cases of metal extensions were

SEPARATION (INCHES)

Figure 45. Effect upon diode voltage of metal
extensions to antenna.

investigated. Figure 45 shows the results. A
12-in. length of pipe changes the voltage about
10 v when connected directly to the end of the
antenna. When pipe and antenna are separated
to leave a i^-in. air gap the effect is reduced to
about 1 v, the order of magnitude of the meas-
ured effect from the flames. If a resonant length
of pipe is used for an extension, the effect is
not so sensitive to separation. Still there is a

SECRET

EVALUATION OF C

marked reduction for an inch separation, which
is about the space observed between the flame
and the projectile. There is, of course, no way
of knowing what the length of a particular flame
may be from any particular projectile. All we
can say from these experiments is that the
effect of flames is consistent with a theory of
antenna length changes.

We see from the figures that it is possible for
a change in antenna length either to increase
or to decrease the voltage by selecting the fre-
quency or length properly. In particular, for

Figure 46. Horizontal differential antenna with
its image.

one assigned length of extension, there is a fre-
quency for which the voltage change due to the
presence of the extension will be zero.

In the course of field tests on the afterburn-
ing problem, fuzes, operating above and below
the resonant frequency, were tried to see if the
effect of the flame could be reduced. No im-
provement resulted, presumably because (1)
the flame lengths were too variable, or (2) the
transient setup by its change of length or posi-
tion was too great.

The discussion of the response to flames has
so far been concerned with the magnitude of
the effect. There remains the problem of its
time dependence. If the changes of ZX1 set up
by the afterburning flame have the same time
dependence as the expected signal, no internal
circuit can discriminate against it. (This as-
sumes the use of a load-sensitive r-f system.)

To date it has been impossible to learn much
about the wave form of the afterburning signal.
Static measurements leave out the large effects

of the airstream on the flame behavior and
hence give only crude qualitative answers.

Some investigations have been made to see
if a change in the frequency response of the
audio-frequency control circuit would reduce
the response to afterburning. No significant
results were observed. This probably means
that the actual afterburning wave form is so
erratic that it has sizable components in the
pass band of an otherwise acceptable control
circuit.

Real improvement has, however, been made
in reducing the afterburning effect by chang-
ing the design of the propellant or the motor
or both.10 These are temporary expedients, since
it is unwise to have the fuze properties dictate
propellant and motor design.

There are definite lines of attack which indi-
cate that the response to afterburning and even
main burning may be reduced to a negligible
value by changing the basic design of the fuze.
However, a discussion of such changes is be-
yond the scope of this volume.

2 14 EVALUATION OF C

This section23*24 and the following contain
additional material supplementing some of the
discussions in the preceding section.

In Section 2.4 two relations were derived
which are here repeated for convenience.

V(Z oR.G/4*) mt)

Zl3 = -^^\/RsiRssGiG3 fi(dn,<f>i3) f 3(631, (j)3i).

(cos r)je -2,rr/x) . (42)

It has been shown that C is a constant for all
antennas. The evaluation of C allows the com-
pletion of the general equation (42) for the
mutual impedance between any two antennas
(only radiation fields considered).

Since C is a constant, we are justified in
choosing the simplest possible antenna in the
evaluation of C. The antenna chosen is the
infinitesimal or differential antenna. For such
an antenna, f(6,<\>) = sin 6 and G = %. The
current, whose absolute value will be called 70,

THE RADIATION INTERACTION SYSTEM

is constant over the infinitesimal length of the
antenna. The power W radiated is

W = ^£. (153)

Any change A Rs in the radiation resistance is
accompanied by a char ge A IF in the power radi-
ated, given by

current 70 in No. 1 constant in the presence of
No. 2, we in effect cancel out the scattered wave
from No. 1, since the scattered wave results
from an additional current in No. 1. Thus we
need consider only the direct contributions from
No. 1 and No. 2. At any point P on the hemis-
phere S, whose radius is R (Figure 46), the in-
stantaneous electric field due to No. 1 is

AW =

I o2A R,

(154)

Ei = sin 6 e&* ™ *)]. (155)

The above expression is based on the assump-
tion that the current remains constant.

These relations will be used in evaluating C.
We shall obtain the mutual impedance between
an infinitesimal antenna and its image in the
following manner. First we shall compute the
additional power AW, which the antenna has
to radiate to maintain its current 70 constant
in the presence of the image. Then, utilizing
equation (154), we shall obtain the resistance
component of the mutual impedance. Finally,
with the aid of equation (42) C will be found.

Consider the differential antenna to be placed
horizontally at a height h above an infinite
perfectly conducting horizontal ground (Fig-
ure 46) . The antenna and its image, to be called
respectively No. 1 and No. 2, form a system
of two interacting antennas. For this system
fi(0i2f<t>i2) — f 2(^21, 4>2i) — 1> and cos t = 1.

We now have to compute the additional power
radiated by the antenna to maintain its free-
space current 70 in the presence of the ground
or image. At this point one advantage of using
the infinitesimal antenna in this calculation
may be mentioned. It is only for such an an-
tenna that we can be sure that the free-space
current distribution can be maintained in the
presence of the ground.

To find AW we can integrate the Poynting
vector over a sphere S of large radius. This
gives us the total power ; subtracting the free-
space power W0, we obtain AW.

The fields along the surface of the sphere are
due to contributions from the real antenna and
its image. In general, some of the radiation
from the image (antenna No. 2) is scattered
by the real antenna, No. 1. The effect of No. 2
over the surface S is the sum of radiation from
No. 2 plus scattering from No. 1. By holding the

The field due to No. 2 is

Ez = — ^ sin 8 e&* -<*<*+'“ *».

In equations (155) and (156)

7 _ \ZqRsG
k -

(156)

\[/ = the angle which the radius vector R
makes with the vertical (Figure 46).

/3 = 27r/X.

The other symbols have been previously defined.
Inequations (155) and (156) the induction and
quasi-static fields have been ignored, since they
do not contribute to the power radiated. Adding,
we have

E = E\ -(- E2

= ^ sin 6 e^031
n

- PR)

jgi/3/i cos t _ e - jph cos (157)

To find the total power W, we have to integrate
\E\2/Z0 over the hemisphere. The details of the
integration will be omitted. The result is

w 47r/o2/c2 , 7r7 o2k2
W = —^7? 1 ~ — •

3Z0
sin 2 (3h

2 W

cos 2j Qh +

4 m

The free-space power W0 is
4tt70 2fc2

Wo =

3Z(

3 sin 2/3/i J.

(158)

(159)

Therefore,

IF — IF 0 A IF ARS _

W0 Wo Rs

1 [w sin m ~ 2<m cos 2/3/1 + w sin wh]

(160)

PATTERN ERRORS DUE TO GROUND REFLECTION

It will be noted that as h approaches zero
(A Rs/Rs) approaches —1, showing that the an-
tenna does not radiate when on the ground, as
is known.

If h be selected large enough so that the
inverse square and cube terms in h may be neg-
lected, we have

ARS 3X . Mi

Rs ~ 8irh Sm X '

(161)

Thus we see that under such conditions the re-
sistive component of the reflected impedance is
proportional to Rs and varies harmonically with
the separation between antennas (results ob-
tained previously by other means). With the
aid of equation (42), it is now seen that

(2) a reflected ray from the ground. It is con-
venient in computing these effects to treat the
radiating system as consisting of the transmit-
ting antenna and its image.

In Figure 47 the coordinate system is a rec-
tangular xyz system with origin at the center
of the image antenna; the image antenna lies
along the y- axis. The real antenna is situated
at a height z = 2h above the xy plane. The
receiving dipole is at a distance a from the
transmitter and is always tangent to a circle
of radius a whose center is the transmitter.
The angle 6 is the angle of azimuth and repre-
sents the angle of rotation of the antenna in
the field setup. The angle of elevation of the

CZp „ r _ 3A „

8t rh ~ 8t rh 8

Since G = •%, we obtain

(162)

C = 9?. (163)

It is also interesting to note that this calcu-
lation, which is based upon radiation fields
alone, shows the contribution to the radiation
resistance arising from the interaction between
inverse square and inverse cube fields of an-
tenna and image, which of themselves do not
radiate power on the average.23’ 24 While this
argument holds for the infinitesimal dipole, it
does not hold for large antennas, since the prox-
imity to ground may alter the current distribu-
tion on the finite antenna. The argument gives
correct results for distances large compared to
the dimensions of the antenna.

2.i5 PATTERN ERRORS DUE TO
GROUND REFLECTION

The experimental setup for measuring direc-
tivity patterns has been described in Section
2.8. There are certain errors inherent in these
measurements because of the reflection from
the ground ; it is the purpose of this section to
discuss these errors.

Referring to the field setup described in Sec-
tion 2.8, the resultant field strength at the re-
ceiving antenna is composed of two parts : (1) a
direct ray from the transmitting antenna, and

Figure 47. Coordinate system used for com-
puting reflected field in radiation pattern setup.

dipole with respect to the image antenna is a.
The distance from the image to the dipole is r.
It is seen that cos a = (a/r).

The ray from the real antenna to the dipole
makes the angle 6 with the axis of the antenna
and is perpendicular to the dipole. The electric
vector associated with this ray is parallel to
the dipole. The ray from the image to the dipole
is also perpendicular to the dipole and makes
an angle y with the image. The term y, as shown
in the diagram, is given by the relation

a cos 6

cos y = — - — = cos a cos 6. (164)

From the relation of equation (164) , y is plotted
versus 6 (Figure 48) for a = 15 degrees, which
represents the actual situation for the measure-
ment of a large number of patterns. Now the
electric vector associated with the reflected ray

SECR

THE RADIATION INTERACTION SYSTEM

is not in general parallel to the dipole. The com-
ponent Er of the electric vector of the reflected
ray, parallel to the dipole is given by

Er = E( 7) cos r. (165)

In the above equation E ( y) represents the radi-
ation field in the direction y from the image
antenna. The term t is the angle between the
direction of E( y) and the dipole. The term
E( y) is in the plane determined by r and the y
axis, and is perpendicular to r.

Then t may be evaluated as follows : The di-
rection cosines l, m, and n, of the dipole are seen
from Figure 47 to be

l = cos 6,
m = sin d,
n = 0.

For the direction of E( y) we obtain the corre-
sponding values :

_ a sin 6 _ cos a sin 6
r tan y tan 7 ;

m = sin 7.

Then

cos 0 cos a sin 0 . . _ .

cos r = + sin 0 sin 7,

tan 7

cos 7 sin 0
tan 7

sin 0
sin 7*

+ sin 0 sin 7,

Thus we have

Er = E{ 7)

sin 7

(166)

The correction factor (sin 0/sin y) = cos t is
plotted versus 0 in Figure 48 for y = 15 de-
grees.

The total field strength parallel to the receiver
is the resultant of Er and the field associated
with the direct ray, denoted by Ed. This re-
sultant we shall call Er

We may write Ed and Er as follows :

4Mei m

- (3a -e (0)]

(167)

Er =
Also

nAf(y} sin 8 ^ _ * _ <w _ „
r sin 7

(168)

Et = Ed+ Er . (169)

In the above equations, A is a constant of

proportionality, and f(0) and / (y) are the
magnitudes in the true radiation pattern for
6 and y; the terms e(0) and e(y) represent
the phase of the radiation field for 0 and y;
a and r are the distances from the receiving
dipole to the fuze antenna and to its image
respectively ; n is the magnitude of the

Figure 48. 7 and cos r versus 6.

reflection coefficient, appropriate to the type
of ground under consideration ; $ is the
phase angle associated with this reflection ;
P= (2:tA).

The image representation used here does not
fully represent the changes in polarization oc-
curring at reflection when n 7^ 1. An examina-
tion of the geometry shows that the receiver
dipole responds only to the component of the
electric field that is parallel to the ground at
reflection. The vertical component whose ab-
sorption is most sensitive to ground properties
is ignored. Thus we may safely use the image
representation with an effective reflection co-
efficient n. It may be noted also that this coeffi-
cient is a constant for all angles 6 of the trans-
mitter, since the angle of reflection to the re-
ceiver is not altered.

A square law detector is used, so that the

SECRET

PATTERN ERRORS DUE TO GROUND REFLECTION

measured directivity pattern, when normalized
to unity, is given by

\Et(e) [2

I ’max’

where |Z^|max is the maximum value of | Et\.
What is desired, however, is the true pattern
given by

\Me) |

-Pie).

Now n/r may be written as N/a, where N, as

Figure 49. f(d) and f(y) cos r for electrically
long antenna.

where A' is a new constant of proportionality.
Thus

jf#n - ™

where q is the expression in brackets in equa-
tion (171).

When the fuze antenna has the pattern of an
elementary dipole, we have

e°

Figure 50. Typical theoretical pattern /2(0),
with per cent error due to ground reflection.

thus defined, differs in general by a few per
cent from n. We may then write

E t{6) = — ej[ut - &a - «(*)].

sin

f(6) + NfM e’T* W +£ W1
sin 7

where we have defined

$i = —(3(r — a) — 4>.

(170)

Then

\Et{e)V = A’ [

m + N^y)8^

+. mw(y) gjpj cos (*, + €(«)

c(t)
(171)

which makes

f(y) ^ = m. (174)

sin 7 v J

Furthermore, there is no phase dependence on
6 ; that is

t{6) — e(v) = 0.

(175)

Then

q = S\d) [1 + N* + 2N cos $J,

(176)

so that

\Etm . m rfn)

1 Ei | 2max " PWrr** 3 W‘

(177)

Thus for this limiting case, the errors reduce
to zero.

SECRET

THE RADIATION INTERACTION SYSTEM

We have seen that when the conditions of
equations (174) and (175) hold the errors re-
duce to zero. In practice, for the fuze antennas
in use, these two conditions are so nearly ful-
filled that the errors are small. Figure 49 is an
example of how nearly equation (174) is ful-
filled, even in the case of complicated patterns.
The solid line represents a three-lobed pattern
f(6) obtained with an electrically long projec-
tile. Although this is an observed pattern, it
represents a possible true pattern. The crosses
are the values of /(y) cos t obtained by utilizing
the relations in Figure 48. For simpler patterns
the differences between f{6) and /( y) cos t are
less than those in Figure 49.

It was mentioned in Section 2.8.2 that the-
oretical patterns representing very good fits for
the observed patterns may be obtained by as-
suming a current distribution of the form
shown in equation (98). Such computations
afford a means of estimating e(<9) — e(y) and

also /( y) cos x. Such calculations over a range
of conditions indicate that s (0) — e(y) does
not exceed 5 degrees for patterns now in use. By
using the values thus found and combining
them with a range of assumed values for 4>,
and n, values of q/qmax may be computed. Such
computations lead to the conclusion that the
error from those sources will rarely be in excess
of 5 per cent.

Figure 50 is an illustration of the extent of
the errors found by such considerations. The
solid line represents the /2(<9) calculated from
the type of current distribution mentioned, with
R and 5 chosen to give a typical bomb radiation
pattern. The dashed curve represents the per-
centage of error obtained by assuming <I> =
(jt/2), n = 1, values which give approximately
maximum errors. The percentage of error curve
is a plot of

™[q/rmmm] versus e-

Chapter 3

ELECTRONIC CONTROL SYSTEMS5

The basic physical phenomena underlying
the production of an actuating signal for a
doppler-type radio fuze have been discussed.
This chapter is concerned with the problems
of designing electric circuits to convert the
signal so that a missile will be detonated in
accordance with the military requirements. In
the preceding chapter it was shown that the
interaction between a radiating system and a
reflecting target can be considered as a load
variation across the two terminals connecting
the antenna with the oscillator. The variations
in load occur at an audio rate. The problem of
this chapter is to show how the variations in
antenna load are converted to a signal which
will detonate the missile at the proper point on
its trajectory.

There are five major subdivisions in this
chapter :

1. The r-f section which treats of the design
of oscillator detector circuits which respond
properly to variations in loading.

2. The audio-frequency section which dis-
cusses methods of controlling the load-variation
signal so that it will reach the proper amplitude
at the proper time.

3. The detonator section in which it is shown
how an audio signal of requisite amplitude ini-
tiates an explosive train.

4. The power supply section in which ways
and means of supplying electric energy to the
electronic circuits are described.

5. A coordination section in which the vari-
ous design compromises are discussed.

31 RADIO-FREQUENCY SYSTEMS*

311 General Requirements of the R-F Unit

The r-f system was originally conceived as

a This chapter, which consists of five major sections,
was prepared by several different authors. They are
named in footnotes to the headings of the various
sections.

b This section was prepared by Chester H. Page, of
the Ordnance Development Division of the National
Bureau of Standards.

a combined transmitter-receiver, converting the
target-approach doppler frequency into an
audio-frequency signal by rectification. The cir-
cuit engineering is simplified by viewing the
net electromagnetic behavior of the radiating
missile as a two-terminal variable impedance.
For practical purposes, it is sufficient to con-
sider this impedance as the parallel combina-
tion of a constant reactance and a variable
radiation resistance. The fixed reactance branch
can be mentally combined with the transmitter
circuit, simplifying the problem to that of an
oscillator feeding a variable resistance load.

The net radiation resistance load is a func-
tion of fuze and missile dimensions as well as
operating frequency. The fuze and missile com-
binations in use lead to radiation loads ranging
from 1,500 to 150,000 ohms, a total range of
two decades. In general, the low end of this
range is associated with long missiles, such as
the larger rockets and bombs; the medium
range (up to 20,000 ohms) is associated with
medium size bombs ; and the upper range with
the small mortar shells. The extreme case of

150.000 ohms is contributed by the fuzes using
transverse-dipole or loop antennas. The small
mortars present radiation resistances from

6.000 to 100,000 ohms, by virtue of the extreme
frequency range used.

The most severe aspect of the large load
range is its effect on the design of a “universal”
fuze. A fuze designed for interchangeable use
on all bombs must operate satisfactorily over
at least a tenfold range of values of load re-
sistance. When the use of a fuze is limited to
a specific missile, the circuits can be designed
for the optimum match between source imped-
ance and load. The goal of semiuniversality, to
reduce the required number of models, places
a severe limitation on the types of r-f systems
that can be employed.

The most elementary r-f system for a prox-
imity fuze consists of a low-power oscillator,
relatively heavily loaded. This may be consid-
ered to be an “oscillating detector” and is oper-
ated under approximately the same condition

ELECTRONIC CONTROL SYSTEMS

as utilized for autodyne reception of telegraphic
communications. The basic design consists of
an oscillator with little regeneration, operating
under Class A grid conditions, developing its
own grid bias across a large grid leak resistor
(of the order of a megohm). The plate current
is supplied through a resistor of some 50,000
ohms. The coupling between the oscillator and
antenna is sufficiently tight to place the oscil-
lator on the verge of instability from overload.
Under these conditions the plate current (and
therefore the plate potential also) is a sensitive
function of load resistance. Such a scheme al-
lows the conversion of radiation resistance
variation into an audio signal appearing across
the triode plate circuit resistor. The funda-
mental weakness of this circuit arrangement
lies in the small range of radiation load for
satisfactory operation. This precludes its use in
semi-universal fuzes, and also leads to critical
load coupling adjustment. Little attention has
been paid to this type of circuit.

Another circuit based on the concept of sep-
arate functions of transmission and reception
used a stable power oscillator inductively
coupled to the antenna circuit. A tuned diode
detector was also coupled to the antenna circuit
for rectification of the doppler frequency
beats. Very early in the program, it was
realized that the transmission-reception-detec-
tion problem could be considered as a vari-
able antenna resistance problem, as previously
discussed. This realization led to a simplifica-
tion of the circuit, by combining the tuned diode
circuit and antenna coupling functions. The
new arrangement comprised a tuned diode volt-
meter across the antenna terminals, with the
diode-antenna tuning coil inductively coupled
to a stable oscillator operated at full power.
This arrangement required an adjustable ver-
nier tuning capacitance for individually res-
onating the diode-antenna circuit to the par-
ticular oscillator assembly. Aside from produc-
tion problems and effects of aging on the tuned
circuit, this design leads to difficulty for semi-
universal application by virtue of the different
antenna reactance presented by different mis-
siles. Although this reactance variation for one
family of missiles is not large, it is sufficient to
produce appreciable detuning of the sharply
resonant diode circuit.

A further simplification of the fuze was based
on the dependence of grid voltage of an oscil-
lator on its load. Details of a practicable circuit
were worked out in cooperation with Andrew
Stratton of the British Ministry of Aircraft
Production during an extended visit to the
National Bureau of Standards [NBS].79’88 If
the oscillator is operated under appropriate
conditions of grid current and grid bias, its
plate current is insensitive to load, but its grid
bias exhibits a smooth reproducible dependence
on load. This is, of course, a variable efficiency
oscillator. The bias developed is almost exactly
proportional to the voltage developed across the
antenna. The antenna is tightly coupled to
the oscillator, and the lack of sharply resonant
coupling circuits makes the system insensitive
to small antenna reactance differences. For the
same reasons, operation is not sensitive to fre-
quency differences among individual oscillators,
and no vernier tuning adjustment need be made.
This so-called reaction grid detector [RGD]
circuit was used in all the later models of prox-
imity fuzes developed by Division 4.

A second type of oscillator reaction which
can accommodate a wide load range was devel-
oped and employed by the Westinghouse Elec-
tric Corporation.111- 205) 206 This circuit is super-
ficially the original oscillating detector with
the plate resistor replaced by the primary
winding of an audio transformer. It differs in
the operating conditions of the triode. The
plate current and generated power are consid-
erably higher than in the oscillating detector,
but the variation of plate current with load is
still employed as the signal generating means.
This circuit is referred to as the power oscillat-
ing detector [POD]. The signal voltage gen-
erated by the load resistance variation is the
equivalent plate circuit voltage which would
produce the observed current variations through
the transformer impedance and triode plate re-
sistance. The grid operates under Class A con-
ditions, instead of the heavy Class C condition
utilized in the RGD circuit.

3,1,2 Sensitivity

Definition of Sensitivity

One fuze will be called more sensitive than
another fuze if it will function further from

SECRET

RADIO-FREQUENCY SYSTEMS

the target, all conditions of use being the same.
This is a purely qualitative concept, which can
be made quantitative in various ways. For ex-
ample, the “Michigan sensitivity” (see Section
2.11) of a fuze is the theoretical function height
over a perfect reflector of infinite extent with
the missile approaching the target plane in the
most favorable aspect and with the speed ap-
propriate to the most favorable doppler fre-
quency (audio-amplifier response). The func-
tion height under practical conditions is pre-
dictable from the Michigan sensitivity by ratio
computations. This definition is still too gen-
eral for our needs. What is desired is an ab-
solute definition of the sensitivity of the oscil-
lator to radiation load changes as shown in
Section 2.7. This relates a given physical situ-
ation to the audio signal voltage produced by
the oscillator system. The knowledge of this
voltage, together with the known characteris-
tics of the amplifier and thyratron, allow the
prediction of function heights in a straightfor-
ward manner as shown in Section 2.9. The r-f
system acts as a means of converting a physical
electromagnetic situation into an electric cir-
cuit problem. In this work, the unqualified term
“sensitivity” has been restricted to the sensi-
tivity of this converter and has been defined as
the developed signal voltage divided by the frac-
tional change of load resistance resulting in
this signal7 [equation (84) of Chapter 2]. Math-
ematically it is defined for infinitesimal load
changes and is the derivative of the operating
voltage whose changes become the audio signal,
thus:

dR/R'

(1)

Since V is the voltage (grid bias or diode out-
put) at the operating point, and dR/R is di-
mensionless, the sensitivity is expressed in
volts. Rewritten in the following form it is the
same as equation (84) in Chapter 2.

S =

d In R’

(2)

where In R refers to the natural logarithm. This
form of the definition is more useful, since it
shows the sensitivity of the oscillator to be the
slope of its “load curve” plotted on natural
semilog paper.

We are concerned here primarily with sensi-
tivity due to load resistance changes rather than
load reactance changes. The possible effects of
the latter are discussed in Section 3.1.

It has been found that properly designed
oscillator-diode [OD], RGD, and POD systems
behave like ideal generators of fixed internal
resistance, with the d-c operating voltage pro-
portional to the load voltage.111 This idealized
r-f unit is quite amenable to mathematical
analysis, and some interesting general relation-
ships are derivable.

Let us consider the behavior of a constant-
current generator with internal (shunt) re-
sistance Ri and unloaded terminal voltage F «.
The terminal voltage for any load is propor-

Figure 1. Circuit with constant current gen-
erator and shunt load.

tional to the net resistance of the load and Ri
in parallel (see Figure 1). Hence, operation
under load R yields the voltage

y _ V co RRj _ y R

Ri R -\- Ri R Ri

(3)

The sensitivity may be found from equation (1)

S = R

= Fc

RRi

(R + Ri)2

= Foo P(1 - P), (4)

where p is the “loading ratio,” or the ratio of
loaded (operating) voltage to unloaded voltage.

The term V is, of course, the operating volt-
age under radiating conditions, and Vm the
voltage when the fuze is properly shielded so
that the oscillator does not radiate.

The final form of equation (4) shows that
loading to one-half the unloaded voltage yields
the maximum sensitivity for a given oscillator
but that this adjustment is not critical (see
Figure 2). This loading ratio is also the condi-
tion of maximum radiated power, or the con-

SECRET

ELECTRONIC CONTROL SYSTEMS

dition of matching the load to the internal
resistance. The problems involved in obtaining
this match by the use of an impedance trans-
forming network between oscillator and an-
tenna will be discussed later.

A load curve (a plot of operating voltage

Figure 2. Variation of sensitivity S with load-
ing ratio p.

versus the logarithm of the load resistance) for
the ideal generator is shown in Figure 3. It is
seen to be a symmetrical S curve. For purposes
of comparing actual generator performance
with this ideal characteristic, such a curve is
not convenient. The ideal case can, however, be
expressed as a linear relation, allowing easy
evaluation of experimental data. This form is
derived from equation (3) by algebraic manip-
ulation and is

so that a plot of 1/V versus 1/R is a straight
line whose intercepts are 1/Fooand — 1 /R{. This
form is exceedingly convenient for smoothing
experimental data and for determining the in-
ternal resistance (Rf) of an oscillator.

This representation of ideal generator be-
havior allows easy comparison of actual per-
formance data with the idealization. Good RGD
oscillators follow this relation quite well over
the load range for which their plate current is
constant. If the feedback in the oscillator is not
optimum, the plate current will vary with load.
It has been found that in this case IJV is a
linear function of 1/R. Since the grid bias is
normally obtained across a grid resistor with
ground return, it is proportional to grid cur-

rent, and the above relations would have the
same form expressed in terms of grid current
instead of grid bias. In the special case where
the grid resistor is returned to an initial bias,
usually positive, the grid bias and grid current
are no longer proportional, but are linearly re-
lated. Equation (5) is then no longer valid. A
plot of Ip/Vg versus 1/R is concave upward
(for positive initial bias), while a plot of Ip/Ig
is concave downward. A straight line is yielded
by plotting Ip/y/IgVg versus 1/R.

For the normal grid resistor connection, the
result that Ip/Vg is a linear function of 1/R
can be directly interpreted to mean that the
oscillator is a current generator of fixed in-
ternal resistance whose current is proportional
to the triode plate current. The results of the
more complicated case where an initial bias is
used imply that the proportionality between

Figure 3. Loading curve for ideal generator.

grid bias and the fictitious terminal voltage is
not the basic relation but that the general phe-
nomenon is proportionality between grid power
and the square of the terminal voltage. This
covers all the above cases.

These relationships, equation (5) and modi-
fications, are not directly applicable to the POD
oscillator, where the variation of plate current
with load is the signal generating means. Ex-
amination of experimental data for this sys-
tem111 showed the plate current Ip to be a linear
function of 1/R over the load range of interest
(see Figure 4). The equivalent signal voltage
in the plate circuit is readily computable from

RADIO-FREQUENCY SYSTEMS

the total plate circuit resistance, so that the
effect of the transformer primary impedance
at any audio signal frequency can be easily
taken into account. These results are mathe-
matically expressed as

I = Io o + ^

The justification for replacing RpI by the
supply voltage En in the last step is experi-

Figure 4. Loading curve (for POD generator)
plotted against reciprocal load. Relation over
range of interest (i.e., resistance values above
100,000 ohms) is linear. For lower resistance
values, solid line represents actual values,
dashed line represents ideal linear extension.

mental. Measurements of Im versus EB on pro-
duction assemblies showed that the dynamic
plate resistance Rp was equal to the static plate
resistance ( EB/I oo) under the operating condi-
tions of this oscillator.

The direct practical application of all the
above sensitivity formulas is limited by the fact
that the radiation resistance is not a free vari-
able. If it were, it could be chosen to match the
source resistance, and maximum sensitivity and
power radiation would be obtained. The oscil-
lator design problem would then essentially re-
duce to the problem of designing for maximum
grid bias under no load.

The radiation resistance of transverse anten-
nas is restricted to high values by the small
dimensions involved. On the other hand, the
radiation resistance of longitudinally excited
antennas is adjustable through a considerable
range of values by variation of the size of the
exciting end cap. Unfortunately, for a given
overall fuze length, increasing the length of the
end cap involves decreasing the separation be-
tween the end cap and the missile. (The effect
of geometry of the end cap on antenna react-
ance has been shown in Figures 4, 12, and 18
of Chapter 2.) This increases the shunt ca-
pacity presented to the oscillator and decreases
the internal resistance that can be had. There
is obviously some optimum compromise between
radiation resistance and shunt capacity for a
fixed set of oscillator design factors.

For a given cap and oscillator, the use of a
matching network suggests itself. Practically,
the network losses frequently cancel the ex-
pected gain of sensitivity. The general proper-
ties of this phenomenon are readily derivable.
Let us assume an antenna of resistance A con-
nected to a simple generator of voltage E by

Figure 5. Block diagram for matching network
for generator and antenna.

way of a passive four-terminal network, as
shown in Figure 5.

The input resistance of the network is given
by R = V/IG. We are interested in the value of
dR/R for a given dA/A. We note first that

and

die

Ig ’

(7)

The effect of increasing A by a small change dA
is the same as would result from the introduc-
tion of a voltage de — IAdA into the output cir-
cuit. The incremental generator current dIG
produced by de is, by the reciprocity theorem,
the same as the increment dIA that would result
from the introduction of de into the input cir-
cuit. If we summarize certain properties of the

SECRET

ELECTRONIC CONTROL SYSTEMS

particular network in terms of a transfer con-
stant T , so that

I A = TV, dIA = TdV, (8)

we readily find

die = -TIAdA. (9)

To evaluate equation (7) we need an expres-
sion for IG in terms of IA. This is obtained in
terms of the network efficiency. The power input
is VI G, and the power output is Ia2A. Hence, the
power transfer efficiency of the network is
given by

I a2 A I aA fifh

e = m = T 77- (10)

Combining equations (7), (9), and (10) gives

dR _ dA

I'‘T

(11)

so that the power transfer efficiency 8 of the
network is also the sensitivity transfer effi-
ciency.

This result suggests the existence of a gen-
eral relation between sensitivity and radiated
power, the source being unchangeable. We can
generalize the oscillator circuit as comprising
a triode, coupling network, and antenna. Viewed
in this light, the idling bias Vm is determined
by supply voltage and tube design and does not
depend upon circuit losses. We assume through-
out that the grid drive conditions are such that
the tube behaves as a constant-resistance gen-
erator. This implies for the RGD that the plate
current is approximately independent of load.

The generalized circuit is shown in Figure 6,
using the constant-current generator represen-
tation for convenience. The network is char-

Figure 6. Generalized circuit arrangement for
coupling generator and antenna.

acterized by the two parameters T and e, previ-
ously defined. The net effect of the antenna and
network is to present a resistance R to the gen-
erator, as indicated in Figure 1.

We have the following starting point rela-
tions :

RIg
elo

We immediately derive0

73 6

AT 2*

= TV,

= V,

= TIaA,

R T- Ri

(8)

(10)

(3)

(12)

Further, utilizing equation (11), we have

o _ dV _ dV_ RRi

b ~ dA/A ~ edR/R ~ €Vc° (R + Ri f

Now the radiated power is

= ISA = T2V2A = TWJA
so that

(13)

, (14)

S eR i Ri

P~A = T2VmAR = Tj

(15)

and is independent of T , s.

This result is of great interest and is in
agreement with intuition. For given triode op-
erating conditions, the sensitivity is propor-
tional to the power radiated. The components
in the network can be adjusted for maximum
voltage across the load resistance, and the sen-
sitivity will be maximized. The unavoidable
practical interdependence of T and 8 does not
affect the relation between sensitivity and
power.

The above discussion reduced the ideal gen-
erator to the triode itself, with V » essentially
a tube parameter. In practice, the idling bias
Foo is defined as the bias with the radiation
load A removed but all other network compo-
nents untouched. This practical definition is
needed, since the presence of the network is re-
quired for oscillation. From this experimental
viewpoint, the effect of the network transfer
constant is to adjust the internal resistance as
seen by the antenna. The inefficiency of the net-
work is expressible as a fixed loss load shunted

c Equation (12), if differentiated, would imply
( dR/R ) = (dA/A). This procedure is not legitimate,
since both T and e are functions of A.

SECRET

RADIO-FREQUENCY SYSTEMS

across the antenna. This reduces the laboratory
idling bias and also lowers the source resist-
ance as seen by the antenna. It has been found
that attempts to increase the step-up ratio of
the network also increase the losses of the net-
work, so that the optimum circuit arrangement
for high-resistance antennas is a compromise
between high bias and load matching.118’ 121

Experimental Determination of Sensitivity

Throughout the early stages of the develop-
ment, all measurements of sensitivity were
measurements of the combined effects of oscil-
lator and antenna performance. Reference is
being made to the pole-test procedure discussed
in Chapter 2. It suffices to repeat here only
that this is a direct measurement of the signal

laboratory evaluation. Laboratory tests are also
much quicker and much more convenient, espe-
cially when several parameter adjustments are
being compared.

Standard laboratory oscillator testing in-
cludes taking a load curve and measuring the
grid bias for various load values. The values
used form a geometric sequence so that the data
points are uniformly spaced on semilog paper
(see Figure 7). Since standard commercial log
paper is logarithmic to the base ten, the slope
dV/d(\og R) must be multiplied by In 10 =
2.303 to obtain the sensitivity S, which is
dV/d{ In R) . The sensitivity can, however, be
conveniently read from a tangent to the curve
by noting the change of ordinate along the
tangent corresponding to two abscissas whose

Figure 7. Typical data points for experimentally determined loading curve ( Eg ). Curve labeled S'
shows sensitivity or slope of Eg curve.

voltage generated upon approach to ground.
When suitable r-f load resistors became avail-
able, and the radiation resistance had been
measured, the oscillator sensitivity S as deter-
mined in the laboratory was found to check
closely with its value derived from the pole
tests. (The derivation is the reverse of the
process of predicting function height for a
given oscillator sensitivity.) The experimental
difficulties of pole-testing, in combination with
ground screen diffraction effects, generally
make this test method less accurate than the

ratio is e = 2.718. Thus holding a straightedge
tangent to the load curve at the point of interest
and noting the intercepts of the straightedge
with the vertical lines R = 1 and R = 2.7, or
R = 3.7 and R — 10 allows computation of the
sensitivity as the difference of the ordinate
values of these intercepts.

Proximity fuze oscillators, like any trans-
mitters, are tested on dummy radiation loads.
The reaction-type units RGD and POD are in-
sensitive to small load reactance errors and can
be tested on resistor loads. The ultra-high-fre-

SECRET

ELECTRONIC CONTROL SYSTEMS

quency resistors of the x/2- and 1-watt size,
F-l/2 and F-l, manufactured by the Interna-
tional Resistance Corporation, have been found
satisfactory. Although the true values of these
resistors at high frequency are not the same as
the d-c values, the percentage difference does
not vary seriously with resistance value. As
discussed in Chapter 2, this allows the employ-
ment of a self-consistent set of resistors where-
in the unit is only approximately the ohm.
Since radiation resistances are automatically
measured in terms of this same unit, proper
dummy loading and load curves are readily
obtained.

In the case of oscillator-diode type fuze,
wherein the antenna circuit is sharply reso-
nant, the dummy antenna must present not
only the correct resistive component but also
the correct reactive component of impedance.
For various practical reasons, such as keeping
r-f currents out of the power supply leads and
metering leads, it has been found necessary to
shield the fuze exciting cap properly from the
laboratory environment. Enclosure of the fuze
oscillator head in a metal shield box normally
introduces more antenna shunt capacity than is
introduced by mounting the fuze on a missile.
To compensate for this, a low-loss inductor is
made a part of the dummy antenna and
shunted across the fuze to be tuned or meas-
ured. This inductor is designed to parallel reso-
nate the excess capacity introduced by the
shield box. The power loss in the inductor is
compensated by appropriate choice of dummy
load resistor, so that the resistive component
of the inductance combines with the test re-
sistance to present the correct net load.

For test operation of the complete metal
parts assembly of a fuze, the shield box must be
rigid and relatively small. (The shield box for
tuning adjustments forms a 2-ft cube.) Since
the fuze is vigorously vibrated in the final test
chamber to search out microphonic defects, a
directly connected dummy antenna is not
usable. The simulated load is capacitatively
coupled to the exciting cap, and the load imped-
ance is connected between the pickup plate
and the chamber. The inductive and resistive
components of this impedance must be em-
pirically adjusted for proper operation. The

operating grid bias (or diode voltage) is meas-
ured as a quality check, but no actual sensi-
tivity measurement is made. The voltage check
for production consistency is sufficient with a
sampling test for oscillator sensitivity.

The load curve slope determination of sensi-
tivity is an indirect measurement. Several
schemes for direct dynamic measurement of
sensitivity have been proposed. These are all
based on the use of a resistance which varies
sinusoidally at an audio-frequency rate and can,
therefore, be used either for measuring oscil-
lator sensitivity or the overall Michigan sensi-
tivity of the fuze. As all these direct dynamic
methods of measuring sensitivity are necessar-
ily signal simulators, they have already been
discussed in Section 2. 12. 10

One dynamic loading arrangement not in this
category has had laboratory use. It is a device
that allows the load curve to be exhibited on an
oscilloscope, showing the existence of any oscil-
lator instabilities and generally simplifying the
process of investigating component changes.
The desired load curve is voltage versus logar-
ithm of resistance. If the load resistance is
caused to vary exponentially with time, then
time becomes proportional to the logarithm of
the load resistance, and the normal linear time
base of the oscilloscope is appropriate for dis-
play of the voltage. The arrangement used com-
prised the dynamic plate resistance of a triode
for the oscillator load, with an appropriate uni-
directional pulse of exponential decay applied
to the grid of the auxiliary triode. Correct
choice of the fixed bias on the grid relates the
dynamic plate resistance to the added bias in
the desired linear fashion.

Practical Oscillator Design

It has not been found possible to design com-
pletely an RGD oscillator on paper. Certain
adjustments must be empirically determined,
and the associated phenomena are not thor-
oughly understood.

Experience has shown that adjustment of
the oscillator parameters to make the oscillator
behavior approach that of the ideal generator
results in the greatest stability and reproduci-
bility of operation. The feedback is adjusted
by varying the plate circuit inductance to the

SECRET

RADIO-FREQUENCY SYSTEMS

end that the plate current is substantially inde-
pendent of load. There is a considerable range
of grid drive (feedback) that will satisfy this
condition.

Within this suitable range of operating con-
ditions, we find that increasing the drive re-
sults in higher grid bias and plate current with
lower internal resistance. The increase of bias
tends to increase the sensitivity; the decrease
of internal resistance usually tends to lower the
sensitivity. This effect arises in the high-shunt
radiation resistances encountered, leading to an
increase of mismatch with decreased internal
resistance.

These two conflicting factors lead to a rela-
tion between sensitivity and drive which has a
maximum. In practice the oscillators have usu-
ally been designed empirically for this compro-
mise of maximum sensitivity under normal ra-
diation conditions.

The value of the grid leak resistor is opti-
mized quite simply. Variation of the grid leak,
ceteris paribus, results in a parallel variation
of grid bias and an opposite variation in plate
current. Thus larger leak resistances give
higher bias with less plate current, hence also
higher internal resistance, until a plateau is
reached. Still larger resistance values give
negligible improvement, and eventually lead to
squegging. Fortunately, when antisquegg sta-
bilization is used the plateau can be reached
and still allow a safety factor of 2 for stability.

With the NR-3A triodes, 47,000 ohms was
most commonly used for the grid leak. The T-51
fuze used a 33,000-ohm leak, with slightly
higher plate current. (See Figures 14 and 15.)

31,3 Radiating System

The mechanical details of the radiating sys-
tem have been discussed in Chapter 2 along
with the corresponding field patterns and radi-
ation resistance values. In this chapter, the
effect of the choice of radiator upon circuit de-
sign will be discussed.

The original “whip antenna” was basically
a trailing wire, base loaded with inductance.
This presented a relatively low-radiation re-
sistance, and was accordingly series tuned. The

oscillator was inductively coupled to the an-
tenna circuit. A tuned diode circuit was loosely
coupled to a second point in the antenna circuit,
thus providing a means of measuring the an-
tenna current and its variations. This bomb
tail fuze served to prove the feasibility of prox-
imity fuzes and was soon discarded in favor of
an engineered assembly.

The second stage in the development of the
exciting arrangement was the substitution of
a conical cap for the trailing whip. The imped-
ance of this was sufficiently high to permit of
parallel loading, the antenna feed points (cap
and body) being connected across the diode
tuning coil. Early models used a variable series
coupling condenser between the cap and coil.
This was done mainly to allow the vernier tun-
ing adjustment to be made externally, the mov-
able center screw of the simple cylindrical
vernier condenser being threaded through a
nut on the apex of the cap. Corona effects to be
discussed later forced the abandonment of this
scheme. The same arrangement was used in the
first production of MC-382 rocket fuzes. In this
case the corona problem was minor, but the
mechanical instability of the condenser, and the
complications of construction, left much to be
desired. These several problems were solved
by making the tuning adjustment from the
base of the fuze, so that the cap could be direct-
connected to the diode coil, and operate at d-c
ground potential.

Further variations of the end cap resulted in
the antenna ring used on T-50 and related
fuzes. (See Figures 4 and 5 of Chapter 1.)
This design yielded medium radiation resist-
ance values, which were readily matched to the
oscillator, with relatively low fixed shunt ca-
pacity. The ring also acted as a mechanical
guard for the wind vane and as an electric
shield against effects of bearing looseness and
vane end-play.

The latest variation of the end cap is found
in the T-132 and T-171 mortar fuzes. (See Fig-
ure 6 of Chapter 1.) In these designs the cap
has grown in proportions until it is used as the
housing for all components of the fuze except
the oscillator and detonator mechanism. This
makes it feasible to locate a turbo-generator
power supply in the fuze nose.

ELECTRONIC CONTROL SYSTEMS

An interesting antenna problem arose in the
case of the 3-in. antiaircraft rocket. The insu-
lated gap in this case was between the rocket
motor and body, or about one-third of the total
length from one end. The radiation resistance
was of the order of 50 to 100 ohms. The loca-
tion of the insulator required it to be mechani-
cally rugged, and this automatically introduced
high shunt capacity. The final design of insu-
lating “coupler” had 40-ppf shunt capacity.

It had been found with experimental low-
capacity couplers that series feed of the an-
tenna was convenient. The antenna load was
simply inserted into the ground return of the
diode coil. Proper load coupling occurred for a
total antenna shunt capacitance of about 15 ppf.
All attempted designs of coupler not exceeding
this capacity failed to meet mechanical
strength tests, so that attention was turned to
the high-capacity rugged designs.

The final 40-q|if design was incorporated into
the circuit by shunt resonating 25 ppf of its
capacity, leaving in effect a 15-ppf coupler. This
tuning was noncritical and was accomplished
by connecting approximately IV2 in. of heavy
wire across the coupler.

Another exceedingly low-resistance antenna
was encountered in experimental work on tail
fuzes for the 4,000-lb and larger bombs (T-40
and T-43). The tail structure of these bombs is
sufficiently large to be used as a shunt-excited
portion of the antenna, the feed points being
the end of the fin structure and the bomb body.

The remaining antenna structures are those
designed for transverse excitation: the dipole
and the loop. These both present exceedingly
high parallel-radiation resistance, because their
maximum dimensions are so small compared to
the usable wavelengths. From the circuit stand-
point, the loop is ideal. It is used as the plate-
to-grid inductance of a Colpitts oscillator; the
interelectrode capacitances complete the cir-
cuit. It has been found advisable to add ca-
pacity from triode grid to ground to balance
the potential distribution of the loop and mini-
mize longitudinal excitation. The loop is, how-
ever, a very inefficient radiator in such small
dimensions. Its series radiation resistance
varies as the fourth power of its radius, meas-
ured in terms of the wavelength.

Early dipole-exciting circuits used the
dipoles as end loading of the grid-plate Colpitts
oscillator coil. Higher sensitivity was obtained
by inductively coupling a dipole loading coil to
the oscillator coil. The two coils were inter-
wound on a double-threaded form for close
coupling. Maximum sensitivity was obtained by
winding the antenna coil with about one turn
more than the oscillator coil. The radiation re-
sistance presented to the antenna coil is about
140,000 ohms. A convenient sensitivity check
was made by noting the change of grid bias
RGD or plate current POD when a 100,000-ohm
load was presented to an otherwise unloaded
fuze. Theoretically, two load voltage measure-
ments bracketing the operating point are needed
for a sensitivity approximation. (The approxi-
mation involved is that of replacing the tangent
slope by a secant slope.) Both these load re-
sistances must be finite. In practice, all oscil-
lators operating at loads higher than 100,000
ohms can be checked satisfactorily by finding
the voltage drop for the 100,000-ohm load. This
is essentially an empirical measure of the re-
sponse of the oscillator to light loads but can be
justified as a good approximation to equation
(4).

S = Foop(l - p) = Foo(l - P) = Fro - F, (16)
for p = 1, i.e., light loading.

314 Tube Characteristics

General Requirements and Restrictions

Problems arising in the development of tubes
for Division 4 radio proximity fuzes did not
stem from technical considerations alone. Deci-
sions of military policy at staff level introduced
extraneous technical problems of sizable diffi-
culty, as will appear from the following brief
historical review.

In the first successful demonstration of the
radio proximity fuze, February 1941, standard
electronic tubes were used. These were obvi-
ously too large for fuze application and pre-
sented serious microphonic problems. Accord-
ingly, cooperative programs were set up with
Raytheon Production Corporation and the Syl-
vania Electric Products Corporation (then Hy-

RADIO-FREQUENCY SYSTEMS

grade-Syl vania) aiming at the design and pro-
duction of small tubes with the desired electric
and mechanical characteristics. First contacts
were on the usual customer-to-manufacturer
basis and did not involve development con-
tracts. Practically any hearing-aid tube could
withstand the low accelerations involved in the
prospective bomb and rocket applications. The
real problems were (1) reduction of micro-
phony to an order of 30 db better than hereto-
fore realized in the best hearing-aid tubes;
(2) securing of extremely stable and relatively
high-output oscillator performance from the
small-sized tubes involved; and (3) the devel-
opment of suitable diode and thyrafron tubes.

The fuze circuit rendered microphony and
self-noise in the triode oscillator of paramount
importance, with diode and pentode micro-
phonics of next importance and thyratron
microphony of least importance. By the early
summer of 1941, reasonably promising triode
and pentode designs were under way and con-
tractual arrangements had been made with the
two companies and others for continued devel-
opment on all four tube types. Such arrange-
ments were handled through Division A,
NDRC, of which Division 4 (then Section E)
was a part.

Concurrently with this program, Section T,
Division A, NDRC, conducted a parallel devel-
opment program with these and other tube
manufacturers on a similar family of tubes for
the shell-type radio proximity fuze. Here spe-
cial emphasis was placed on tube ruggedness,
with the requirement that a setback of 20,000#
should be successfully withstood. Tube mi-
crophony was apparently not as serious a prob-
lem for Section T use, partly because of a some-
what different lower power oscillator arrange-
ment, but primarily because the centrifugal ac-
tion of the spinning shell tended to keep the
tube element supports in a fixed position.

On August 26, 1941, Dr. Richard C. Tolman,
Chairman of Division A, NDRC, appointed a
committee to coordinate the two tube programs
with A. J. Dempster as chairman, L. Grant
Hector representing Section T, and Harry Dia-
mond representing Division 4, then Section E.
Contractors were informed of this setup. Both
programs were prosecuted in parallel with Sec-

tion T emphasis on ruggedness and Section E
emphasis on microphony and oscillator per-
formance. It is of interest to note that elements
of design introduced to make a tube nonmicro-
phonic go a long way toward making the tube
rugged. The correlation is by no means 1-to-l
but, curiously, the reverse is not nearly so true,
i.e., making a tube rugged does not insure free-
dom from microphony.

As will appear from the following more tech-
nical discussions, many of the expedients for
making tubes nonmicrophonic, such as special
filament tension springs, four-pillar base con-
struction, etc., were known to the art but were
also essential in the Section T program for
making tubes rugged. High-level policy re-
quired that Section E tubes be designed so that
in the event of prior compromise they would
not reveal details of rugged tube design to the
enemy. This policy was based on firm military
considerations and was followed in good faith.
However, it placed the Section E tube program
in the anomalous position of having no recourse
to certain technical expedients known to be
available to the enemy.

Hence, up to the time the mortar fuze design
was begun, problems of Section E tube design
consisted of how to attain the desired electric
and mechanical performance without making
the tubes too rugged. Since maximum rocket
setback was of the order of 400# and some
safety factor was essential, it was specified that
tubes should withstand 2,500# as a lower limit,
but under no circumstances should such tubes
withstand more than 10,000#. The curious situ-
ation ensued wherein anything that made a
tube ‘‘not too rugged’’ was greeted with delight
and tested with the hope that it would not affect
tube microphony. One exception, a GE micro-
thyratron, simulating lighthouse tube construc-
tion, was permitted by common consent, since
it was not used in the Section T fuzes and no
expedient could be found whereby it would not
withstand 20,000 to 30,000#.

In addition to the general requirement that
the tubes fail at high accelerations, the follow-
ing types of structure (most of which were well
known to the art) could not be used: (1) four-
pillar construction for supporting grid and
plate elements, (2) a coil spring cantilever

ELECTRONIC CONTROL SYSTEMS

(mousetrap construction) for supporting the
filament under proper tension, (3) cross-
press construction for the lead end of the tube,
and (4) grid sleeves and grid stops. (See refer-
ence 33 of Chapter 1.)

effect on the feedback. The circuit can be con-
sidered as a Hartley with additional capaci-
tance across the plate and grid coils or, equally,

Triodes

The design starting points were the sub-
miniature hearing-aid amplifier pentodes al-
ready in existence. Omission of the screen and
suppressor grids and replacement of the fila-
ment by a more powerful one made a triode
suitable for experimentation. The power re-
quirements on the triode were so relatively
heavy that Raytheon put in two filament
strands to obtain the desired emission and life.

The final design of the Raytheon tube was
designated NR-3A and has approximately the
following characteristics.

Filament voltage
Filament current
Amplification factor
Mutual conductance
Cutoff bias

1.4 v nominal
220 ma

1,600 micromhos | at -7'5 v bias
—23 v

The above data were obtained at the nominal
plate voltage of 140.

A photograph of the NR-3A triode is shown
in Figure 8 and of the subassembly of the same
tube in Figure 9.

The Sylvania triode NS-3, which was used in
MC-382 battery fuzes but not in generator-
powered fuzes, has approximately the following
characteristics at nominal plate voltage of 140.

Filament voltage 1.4 v nominal

Filament current 140 ma

Amplification factor 9.3 / , .

Mutual conductance 1,350 micromhos )at ' v ias

Cutoff bias — 15 v

This tube was not used in bomb fuzes be-
cause of its low microphonic stability. This
point will be discussed later.

These triodes work well in any of the stand-
ard oscillator circuits. The oscillator-diode type
fuzes used the quasi-Hartley circuit shown in
Figure 10. If the grid-filament and plate-fila-
ment interelectrode capacities were negligible,
this would be a Hartley oscillator using the
grid-plate capacitance as the “tank,, condenser.
In practice the first two capacitances are of the
same order of magnitude as the last, so that the
interelectrode capacitances have considerable

Figure 8. NR-3A triode (left) and NR-2 diode
(right). Arrows show crimps to support mica
spacers. Scale shown is 1 in.

as a Colpitts with an added coil tap. If the in-
ductive feedback ratio is not equal to the capa-
citative feedback ratio, local circulating cur-

rents are created in the grid and plate branches
of the circuit, introducing extra power losses
and sometimes critical response to coil adjust-

RADIO-FREQUENCY SYSTEMS

ments. This circuit operated satisfactorily in
the 120- to 140-mc range, but was unsatisfac-
tory at 150 me.

Later circuits, RGD and POD, used pure Col-
pitts connections (cf. Figures 11 and 12).
These perform quite uniformly over the whole
range of 50 to 200 me that has been used. For
stable efficient operation, the NR-3A triode re-
quires driving to approximately 2-ma average
grid current. That is, in the oscillator-diode
arrangement, where a low-impedance power
source was required, the maximum usable grid
leak was 15,000 ohms and the minimum bias
for proper operation was 30 v, corresponding
to 2-ma direct grid current. In practice, this

Figure 10. Typical quasi-Hartley oscillator used
in oscillator-diode fuze circuits.

current fell between 2 and 3 ma. In the RGD
oscillator, grid leaks of 33,000 and 47,000 ohms
have been used with the idling bias in the range
60 to 100 v, so that the grid current under no
load conditions was 1.5 to 2 ma. Since the opti-
mum grid drive is affected by many factors,
such as power output, internal resistance of the
oscillator as a generator, stability of oscilla-
tion, and sometimes maximum grid bias, it is
not determinable from any simple theory of the
oscillator. It is, therefore, a purely empirical
observation that, in general, the NR-3A should
operate at about 2-ma grid current (this is for
a nominal plate supply of 140 v) .

The plate current of this triode may be any-

thing in the range 7 to 14 ma, depending on
the oscillator frequency and application. The
subject triode has not been found useful at

Figure 11. Typical Colpitts oscillator used in
RGD fuze circuits.

lower plate current because of power (and sen-
sitivity) requirements. Higher plate current
does not normally occur with optimum oscilla-
tor design, but the average current for actual
fuze designs has been found to be approxi-
mately proportional to oscillator frequency in
a given type of application and circuit.

All electric circuits are in some degree sub-

Figure 12. Typical push-pull Colpitts oscillator
used in POD fuze circuits.

ject to spurious signals. The sources of these
signals range from statistical thermal fluctua-

SECRET

■

ELECTRONIC CONTROL SYSTEMS

tions of resistance to intermittent connections.
In the battery-powered fuzes, the most impor-
tant noise sources were inside the triode. This
electric noise can be classified as self-noise and
microphonics. Self-noise arises without appre-
ciable mechanical stimulus of vibration or
shock. Microphonics refers to those noises
which are mechanically induced.

The early triodes frequently were noisy
(self-noise) due to the presence of charred lint.
The lint had become charred in the baking op-
erations and formed a conducting carbon fila-
ment which eventually would bridge two tube
elements. The electrostatic forces on the lint
were responsible for its short-circuit seeking
habits. Occasionally one end of a lint piece
would firmly adhere to the plate and the other
to the grid, forming a miniature carbon fila-
ment incandescent lamp. In such cases, the lint
would often be more luminous than the cathode.
The lint problem was eliminated by improved
manufacturing techniques. Another source of
noise was electric leakage between leads on the
outside of the glass press. This was traced to an
alloying of the glass and a metallic oxide
formed on the external leads in the pressing
operation. There was one lead which was cut
off next to the glass, since it was merely an
anchor. By postponing the cutting off until
after the seal was made, this wire did not get so
hot and did not burn. The extra length served
to conduct heat away. No more trouble was had
from this source after the new procedure was
established. Interelectrode leakage paths in-
side the envelope, i.e., on the mica spacers, can
also produce noise. This phenomenon is dis-
cussed under the diode noise problem, as it was
not serious in triodes.

The most difficult problem in designing the
triode was the reduction of microphonic effects.
In an oscillator, any variations of either the
low-frequency parameters of the triode of the
interelectrode capacitances produce variations
of the high-frequency output and the developed
grid bias. The microphony problem became
acute with the transition from battery power
to generator power, because the rotating sys-
tem associated with the generator necessarily
produces vibration. In fact, as the missile
changes speed, so does the rotating system, and

the frequency of the mechanical vibrations is
apt to sweep across some resonant frequency of
the tube structure.

The most serious resonance was that of the
electrode assembly as a cantilever spring. Fur-
thermore, if the elements are not tightly cou-
pled at the free end, the plate can vibrate
relative to the grid and filament. If the mica
spacer is sufficiently snug to prevent this, then
the whole assembly is a stiffer cantilever but
can still vibrate with respect to the surround-
ings. Of course, the bending of the structure
will also introduce a small relative motion be-
tween grid and anode. Microphony of this type
was practically eliminated by pressing the glass
envelope in against the mica spacers on both
sides. This is referred to as crimping. Since the
electrode support posts lie along the major di-
ameter of the cross section of the triode, crimp-
ing of the flat sides of the bulb (preventing
motion along the minor diameter) greatly in-
creases the rigidity of the structure. This con-
struction was adopted as standard in the
NR-3A triode and was also introduced into the
diodes as a general precaution, although the
need for it in the latter case was not demon-
strable. The arrows in Figure 8 point to the
crimps on the triode.

The filamentary cathode itself cannot be
made rigid. Its resonance frequencies are kept
well above the audio range by proper tension,
but freak low-frequency disturbances can be
generated. These apparently arise from the
nonlinear phenomena associated with finite vi-
bration amplitudes of the filament. If the fre-
quency of the driving force applied to the fila-
ment is slowly varied, the resulting vibration
amplitude increases according to a normal
resonance curve as the filament resonant fre-
quency is approached. As the amplitude in-
creases, the resonance frequency is changed by
virtue of the finite amplitude. When the driving
frequency passes the moving resonance, the re-
sulting decrease of amplitude moves the reso-
nance back, further decreasing the amplitude.
The net result is a sudden drop of amplitude at
driven frequency to the value predicted by the
simple resonance curve. The sudden change of
average tension excites a transient at the na-
tural frequency which produces a beat with the

SECRET

RADIO-FREQUENCY SYSTEMS

driven frequency. Thus, 100-c beat transients
have been observed in a filament driven at ap-
proximately 5,000 c. The extreme sharpness of
resonance of the filament allows this phenome-
non to occur for slight variations in driving
frequency. The details of the effect have not yet
been investigated mathematically.

Another possible source of microphonics is
associated with low filament tension. The fila-
ment passes through a small hole in the top
mica and then runs to a tension spring. For
various reasons, it is best to pull the filament
against the edge of this hole by placing the
spring off center. With low tension it is con-
ceivable that under vibration or shock the
filament will slip on this edge, producing
noise.

Generally, high filament tension is indicated,
but variations in tension adjustment can lead
to filament breakage. The simple construction
utilizing a cantilever spring is sensitive to pro-
duction variations of spring displacement. This
situation can be improved by the use of a longer
cantilever. The extra length is incorporated by
coiling the cantilever into a horizontal helix,
with the last turn straightened out tangentially.
This spring is made of ribbon. Another spring
design that has been used can be readily de-
scribed as two such springs of wire, one left-
handed and one right-handed, joined by a canti-
lever hairpin for the filament support. This
type of construction is referred to as the mouse-
trap spring and was not used in the NR-3
triodes because it was believed it would make
the tubes too rugged.

The electrode structure must be rugged to
withstand rough handling of the fuze as well as
the high accelerations encountered in mortar
and shell firing. Ruggedness is a simple matter
of structure design, the problems arising in
making a sufficiently rugged assembly as
simply and cheaply as possible, and of such
design as to be readily adaptable to the mass
production techniques of tube construction.
The filamentary cathode is the only element
which cannot be braced and solidly supported,
but its mass is very low. Its ruggedness is in-
creased by shortening it, since its total mass is
thus reduced, but its tensile strength is un-
affected.

Diodes

The major requirements on the diode de-
tector were small size, low filament power, and
reasonably low plate resistance. The low fila-
ment power was requisite to battery-powered
fuzes. With the advent of generator power, it
was found advantageous to increase the diode
filament ruggedness at the expense of addi-
tional heating power ’ by increasing the fila-
ment diameter. The average characteristics of
the final design, Raytheon NR-2A, are

Filament voltage 0.60 v

Filament current 70 ma

Effective plate resistance 50,000 ohms

This apparently high plate resistance is satis-
factory, since the diode (see Figure 8) ordi-
narily works into a 1-megohm load resistance.

At high frequency, and high applied voltage,
the capacitative anode-cathode current is an
appreciable fraction of the normal filament
current and can cause burnout. A more serious
burnout problem was caused by stray induc-
tive coupling between the oscillator and the
diode filament circuit.

There have been occasional indications of
diode microphony, but these have been nebu-
lous. Crimping was adopted, as in the triode,
for a general precaution. The high inverse
voltage on the diode did lead to self-noise prob-
lems, involving leakage paths on the mica elec-
trode spacer. These leakage paths could be
eliminated in most cases by “sparking” the
tube. This consisted of playing a high-fre-
quency discharge over the surface of the tube,
which apparently burned the conducting ma-
terial off the mica. A still more effective remedy
consisted of spraying the mica surface with a
thin coating of Alundum. The resulting rough
surface inhibits the formation of leakage paths.

The major source of leakage was found to be
stray deposits of “getter” material. Redesign
of the getter holder was the final step in elimi-
nating leakage.

315 Spurious Signals and Circuit Stability
Component Noise

Not all noise and microphony arises in the
tubes. Occasionally unstable resistors and con-
densers are found which generate noise in op-

SECRET

ELECTRONIC CONTROL SYSTEMS

eration, but this phenomenon is not sufficiently
frequent to be of concern. Most of the residual
microphony can be traced to poor workman-
ship (or design), involving such factors as in-
securely anchored connecting leads and coil
windings and imperfect metallic contacts in the
mechanical assembly. Insufficient restraint of
the triode envelope often results in severe mi-
crophony because of th£ resulting variation of
capacity, when the triode moves relative to its
surroundings. The major part of all micro-
phony is induced by vibration of the power-
supply generator and associated rotating sys-
tem. Dynamic balance of a one-piece rotating
system has eliminated much of this difficulty.
(See Section 4.6.)

The power supply itself can introduce noise
by supplying a modulated plate voltage to the
oscillator. Noise modulation of the supply volt-
age can arise from irregular axial motion of
the generator magnet (rotor) as well as from
such obvious defects of operation as rubbing
of the rotor on the pole faces and intermittent
rotor-stator contact via stray metallic particles.
Instantaneous fluctuations of rotating speed,
such as can occur through the slack of a shaft
coupler, result in fluctuations of output voltage
if the generator is operating on a nonconstant
portion of its voltage-speed curve. Some noise
has been traced to variation of contact between
rectifier elements, but this is eliminated by a
combination of careful element manufacture
and high stack pressure.

Corona Effects

Early model oscillator-diode fuzes employed
the customary series d-c load resistance on the
diode rectifier. This automatically put the recti-
fied signal on the antenna cap and isolated the
cap from ground by the load resistor, normally
1 megohm. Field experience indicated that
change of bomb potential in flight produced
small corona effects. The high-resistance cap
isolation caused the production of a signal-
voltage input to the amplifier, when the charge
on the bomb plus fuze was redistributed. Field
effects of random function and peculiar carrier
modulation could be reproduced in the labora-
tory under the influence of a 300-kv d-c gen-
erator.

This source of malfunction was completely
eliminated by maintaining the antenna cap at
the same d-c potential as the bomb. This was
accomplished by grounding the cap, as far as
direct current or audio is concerned, through
the antenna coil and using a shunt load on the
diode output. All fuzes since have incorporated
the d-c grounding of the antenna. The effect of
this circuit change was reported as follows:218

The rearrangement of the diode coupling circuit in
the ROB [abbreviation for radio-operated bomb fuze]
showed satisfactory solution of the problem of elimi-
nating operation of the fuze by static voltage dis-
charges. With the previous arrangement, the fuze
would function when placed in the neighborhood of a
-30-kv field; with the new scheme, the fuze withstood
visible corona and other discharges when in the neigh-
borhood of a 300-kv field.

Unstable Oscillation

When attempts are made to increase the
power output sensitivity of an oscillator, un-
stable oscillation conditions are frequently en-
countered. For example, increasing the grid
leak resistance increases the bias, internal re-
sistance, and sensitivity of the oscillator while
decreasing the plate current. If the critical
value of resistance is exceeded, however, the
oscillator becomes unstable. This instability
may be great enough to cause alternate periods
of oscillation or may be mild enough to cause
only a low-percentage modulation of the oscil-
lation amplitude. The first effect is the familiar
intermittent oscillation, often attributed to a
large time constant in the bias circuit. The
whole gamut of instabilities is incorporated
into the term “squegging.”

Operation under intermittent oscillation con-
ditions offers interesting possible advantages
resulting from the high ratio of peak power to
average power. This was investigated to some
extent in connection with battery power to re-
duce the average anode current. A peak voltage
detector, such as the diode in the oscillator-
diode fuze, can “remember’’ the antenna volt-
age from pulse to pulse, and the sensitivity
with an intermittent oscillator is approximately
the same as that with a steady oscillator whose
amplitude is equal to the peak amplitude of the
former. This type of operation is not possible
in the RGD, since the rectified output is also

SECRET

RADIO-FREQUENCY SYSTEMS

the oscillator bias. In fact, loading curves show
that the RGD average bias is very insensitive
when the oscillation is intermittent. Detection
of target approach with the RGD might be
feasible by detecting the change of intermit-
tency period with radiation load. Experiments
by A. Stratton in England (communicated ver-
bally) show that the pulse repetition rate is a
smooth sensitive function of radiation resist-
ance. Investigation of this scheme requires
the development of a variable time-delay reflec-
tion line for a dummy antenna, since for non-
steady signals the effect of target reflection can-
not be replaced by an impedance. The lack of
reflected signal during the first few cycles of
each pulse (while the oscillation is building up)
can well make a fundamental difference be-
tween field performance and loading curves
representing a steady-state condition.

Just as steady oscillation would be fatal to a
fuze designed to operate intermittently, squeg-
ging in any form is likely to be fatal to any
fuze of the present types. It is not the inter-
mittency itself that produces early functions,
since the normal repetition rate is of the order
of 100 kc and so does not affect the amplifier.
Rather, it is the marginal stability of the oscil-
lator that does the damage. For example, varia-
tion of the supply voltages can convert a steady
oscillation into an intermittent one ; the change-
over produces transient pulses which are
passed by the amplifier. Under some threshold
conditions, a sensitive superregenerative oper-
ation can occur, amplifying thermal voltages
and other hiss noises.

In the early unprotected RGD units, margin-
ally high values of grid resistor occasionally
produced the modulation phenomenon of the
second type described above. No mention of
this particular phenomenon has been found in
the literature. Only a qualitative theory has
been evolved.

Intermittent oscillation arises from an un-
stable condition in which the oscillator grid
bias increases until plate current and oscilla-
tions cease. This extreme bias decays exponen-
tially with time at a rate determined by the
product of the grid-leak resistance and the bias
storage capacitance. When the bias decays to
a value at which oscillation will start, the oscil-

lation starts and grows in amplitude until the
bias is again too large for the tube to operate.
This starting and stopping of oscillation re-
peats periodically.

The instability represented by the appear-
ance of self-modulation is fundamentally of the
same nature but of a lesser degree. In this case
the oscillation amplitude and grid bias increase
with time, but, before the tube is rendered in-
operative, a temporary equilibrium between
amplitude and bias is reached. Because of time
lag between a change of amplitude and the
resulting change of bias, this equilibrium is not
stable but represents a condition where the
oscillation amplitude is not sufficient to main-
tain the bias. Both start to decrease and con-
tinue to decrease until a lower temporary
equilibrium is reached. At this low equilib-
rium the oscillation amplitude is more than
sufficient to maintain the bias, so that the bias
and amplitude again increase. The phenomenon
is periodic.

Both types of instability arise because of the
presence of an operating point (combination
of grid bias and oscillation amplitude) that
represents an unstable equilibrium. An un-
stable equilibrium is an equilibrium condition
in which any small deviation of the operating
point produces conditions that force the operat-
ing point still further from equilibrium. If no
restoring force is encountered by the operating
point, intermittent oscillations result. If suffi-
cient restoring force is encountered on both
sides of the unstable equilibrium point, the
operating point will oscillate over a range. If
the inherent instability is increased, that is, by
increasing the grid leak resistance, the range
of operating point variation will increase and
finally intermittent oscillation will result.

The stability of the original operating point
depends on the relation between oscillation am-
plitude and grid bias and on the time lag with
which the bias variation follows a correspond-
ing amplitude variation. An operating point is
statically stable if a small arbitrary change
of oscillation amplitude produces a greater
change of bias than would be needed to keep the
bias in equilibrium with the amplitude. A stat-
ically stable operating point will be dynamically
unstable if the bias change does not occur rap-

SECRET

ELECTRONIC CONTROL SYSTEMS

idly enough. This dynamic instability can be
produced by the use of too large a time con-
stant in the grid-bias circuit.

The dynamic stability of the RGD oscilla-
tors has been increased by a circuit whose
essential elements are shown in Figure 11. The
resistor Rp in conjunction with the condenser C
comprises a means of reducing dynamic insta-
bility by obtaining a voltage increment from
the rate of change of amplitude and introduc-
ing this voltage increment onto the grid in such
a manner as to make the bias anticipate any
change of amplitude and prevent its occur-
rence.

The voltage drop across the stabilizing re-
sistor Rp is proportional to the anode current.
For small variations of oscillation amplitude,
the anode current variation is proportional to
the amplitude variation. Hence, the voltage
drop across the resistor has a time rate of
change proportional to the rate of change of
oscillation amplitude. The terminal of this re-
sistor nearer the anode is connected by a small
capacitor to that terminal of the grid leak re-
sistor Rg which is nearer the grid. This capaci-
tative coupling between the stabilizing resistor
and the grid leak causes an incremental volt-
age, which is approximately proportional to the
rate of change of amplitude, to appear across
the grid leak.

Static stability of an oscillator is normally
achieved by the self-biasing action of a grid
leak. If the amplitude of oscillation actually in-
creases, the bias is increased, producing sta-
bility. This stabilizing circuit achieves dynamic
stability by increasing the bias, if the oscillation
amplitude starts to increase. Thus the bias is
corrected if the amplitude has only a rate of
change, without waiting for the change to actu-
ally occur. This means that if the amplitude
starts to change, the grid bias anticipates the
change from the fact that it started and pre-
vents the actual change from occurring. This
anticipation of a change is the antithesis of the
ordinary time lag with which the bias follows
an amplitude change.127a

Antimicrophony Circuits

The audio-frequency signal in a proximity
fuze is produced by detection of a slightly

modulated high-frequency oscillation. The
problems of microphony are intimately associ-
ated with this low degree of modulation, which
is normally about %0 of 1 per cent. This im-
plies that accidental variations of the steady
diode voltage, oscillator bias, or plate current
(according to the fuze type) need be only a
fraction of a per cent in magnitude to generate
spurious signals as large as normal firing sig-
nals. Highly selective amplifiers are used to
discriminate against these microphonic volt-
ages (see Section 3.2). Various schemes have
been proposed to alleviate the situation, and
these are all designed to neutralize essentially
the steady voltage and thereby increase the
fractional modulation produced by a target re-
flection.

In the oscillator-diode fuze, futile attempts
were made to neutralize the steady high fre-
quency applied to the diode. One such sugges-
tion was to arrange the antenna and detector in
a bridge circuit, so that the antenna load varia-
tion would appear as a bridge unbalance. No
workable arrangement has been devised. An-
other scheme applicable only to the oscillator-
diode fuze was based on the fact that oscillator
microphonics produce almost identical signals
on the oscillator grid bias and diode output.
These can be balanced against each other in
a push-pull transformer coupling arrangement.
This worked in the laboratory but would not
in practice because of the exacting require-
ments on tuning accuracy and equality of d-c
grid bias and diode output.

In the RGD, where the signal appears on the
oscillator bias, a simple means is available for
reducing the response to plate-supply voltage
variations. The grid leak may be returned to
the plate supply, instead of to ground, if its
resistance value is appropriately increased so
that the same grid current will result in the
same grid bias. This can be done for only one
operating condition, since the bias is no longer
proportional to the grid current. There will be
a point on this resistor which will be at ground
potential, since the grid end is negative and the
plate end positive. Since the RGD oscillators
are sufficiently linear to develop a bias propor-
tional to the plate-supply voltages over a wide
range, the cold point on the resistor will re-

RADIO-FREQUENCY SYSTEMS

main cold if the plate supply varies. On the
other hand, variations of antenna load, which
vary the bias but obviously do not affect the
power supply appreciably, will generate a volt-
age at the initially cold point. It is apparent
that this tapped resistor is a voltage divider
on the signal and reduces it in the ratio
EB/(Eg + Eb), where Eg is the magnitude of
the bias, and EB the plate-supply voltage. Es-
sentially, the same results can be had for pass-
band and higher frequencies by returning the
grid to the plate supply for audio frequencies
only. This eliminates possible difficulties aris-
ing from the application of positive bias while
the cathode is warming up. This modification is
made by using the normal ground return on the
grid leak and coupling the plate supply to the
amplifier input on the other side of the blocking
condenser which isolates the oscillator bias
from the pentode. These arrangements are sat-
isfactory for rejection of power supply noise,
when the oscillator works into a given load,
such as any one missile. Installation of the fuze
on a different missile generally upsets the bal-
ance, as the operating bias is different.

Another scheme has been proposed for the
RGD, but not experimentally investigated. Its
operation is based on the fact that a normal
RGD oscillator draws a plate current which is
independent of load, so that a high audio im-
pedance in the plate circuit would not affect
normal operation. If any audio voltage appear-
ing across the impedance were properly cou-
pled back to the grid, a high degree of degener-
ation (negative feedback) could be introduced
for spurious signals without producing loss of
sensitivity. This is possible because most spuri-
ous signals (microphones or supply fluctua-
tions) generate in-phase variation of grid bias
and plate current.

Arming Pulse

Safety of the fuze is achieved by mechanical
interruption of the powder train as well as in-
terruption of the electric circuit of the detona-
tor. Details are discussed in Section 3.3. Neces-
sarily, the process of arming a fuze involves
completion of the detonator circuit, and this
can conceivably give rise to an arming pulse
which may prematurely fire the fuze.

Stray r-f currents are usually present to
some extent in the power supply leads and,
therefore, couple into the detonator circuit. The
presence of any r-f current in the detonator,
however small, indicates coupling between this
circuit and the oscillator. Closure of this circuit
will, therefore, change the load on the oscilla-
tor. Since the oscillator is very sensitive to load
changes, a firing strength transient can occur
even though the r-f current in the detonator is
apparently negligible. A by-pass condenser
across the detonator will not usually eliminate
this pulse, but a small series choke will.

A related type of pulse occurs in mortar
fuzes of the T-171 and T-132 types. In these
designs, power supply and amplifier are en-
cased in the exciting cap, so that the detonator
firing current must traverse the antenna split.
This requires a choke in one detonator lead;
the ground return is through the antenna coil.
In this arrangement, the choke does not suffi-
ciently isolate the detonator, since the total gap
potential is across the choke. Connecting the
detonator is essentially the same as connecting
the choke across the antenna, and it produces a
strong pulse. Thorough by-pass of the detona-
tor-switch combination would eliminate the
pulse, but complete by-passing at this point is
not always feasible from the production design
standpoint. Circuits have been devised to im-
munize the fuze against this pulse and are de-
scribed in later sections.

Additional Precautions

There are several problems of circuit detail
that have not been discussed above, being too
minor to warrant special headings. A few of
these will be mentioned here as being worthy
of special precaution.

The grid-bias variations are fed into the
audio amplifier. The input circuits of some
amplifiers can present enough shunt capacity at
high frequency to cause squegging in the oscil-
lator. A series isolation resistor of 100,000
ohms is sufficient protection. This resistor, in
conjunction with grid-to-ground by-passing at
the pentode, also helps to reduce stray r-f volt-
ages on the pentode grid. Stray radio frequency
at this point will be rectified, changing the bias
on the pentode and hence changing the gain.

SECRET

100

ELECTRONIC CONTROL SYSTEMS

Stray r-f currents also reach the pentode via
the filament leads. It has been found advisable
to connect a fairly large capacitance, 150 to
250 across the filament supply close to the
triode. These stray currents in the power leads
also produce electric coupling between the
oscillator and moving generator parts. Vari-
able contact between shaft and bearings can
then produce spurious signals. To minimize this
effect, the plate-supply lead is also heavily by-

Figure 13. Oscillator-diode circuit for T-50 fuze,
diagram arranged to correspond to photograph.

passed to ground as it leaves the oscillator
compartment.

3,1,6 Typical Designs

Details of the various fuzes are presented
later in the “catalog” chapter (Chapter 5). The
purpose of this section will be fulfilled by pre-
senting prototype oscillators.

The oscillator-diode type is exemplified by
the T-50; the circuit is shown in Figure 13R
and the component placement in Figure 13A.

The oscillator-diode circuit had several dis-
advantages. An obvious economic and space
disadvantage is in the need for a diode and
associated components. Accurate tuning of the
diode-antenna circuit is a nuisance in produc-
tion, and temperature and aging effects fre-
quently detune the fuzes. The outstanding de-
fect of this circuit is its microphony associated
with frequency variation of the oscillator. Un-
less the diode circuit is tuned exactly, its sharp

TEST point

A is photograph of oscillator block. B is circuit

response makes it a frequency discriminator,
resulting in spurious signals for any micro-
phonic variation of triode capacitances. The
RGD, on the other hand, is relatively broad in
its tuning effects. That is, a given change in an-
tenna capacity or oscillator frequency results
in a change of grid bias which is very small
compared to the corresponding change in diode
voltage in the OD type of fuze. Detailed com-
parisons are presented in the bibliography.51

Fuzes of the RGD type require no individual
oscillator adjustments and are surprisingly
uniform in production. The design used in

SECRET

RADIO-FREQUENCY SYSTEMS

101

the T-50 series of fuzes is illustrated in Fig-
ure 14.

The dipole antenna-type fuze is illustrated by

coupling. Circuit and layout are shown in Fig-
ure 15.

The simplest circuit of all is that used in the

Figure 14. RGD circuit for T-50 fuzes. A is photograph of oscillator block. B is circuit diagram
arranged to correspond to photograph.

T-51. The dipole is inductively coupled to a T-172 mortar fuze with a single-turn loop an-
Colpitts oscillator, the antenna coil being in- tenna. This consists of a simple squegg-stabil-
terwound with the oscillator coil for close ized Colpitts oscillator, using the loop for the

Figure 15. RGD circuit for T-51 fuze. A is photograph of oscillator block. B is circuit diagram arranged
to correspond to photograph.

SECRET

102

ELECTRONIC CONTROL SYSTEMS

circuit inductance. The circuit is shown in Fig-
ure 16.

31,7 Generalization of Sensitivity Concept

The sensitivity of the r-f unit has been de-
fined as S = R(dV/dR) . This is sufficient for
practical purposes, where the antenna circuit
is operated at resonance. If, however, the an-
tenna circuit is nonresonant or if reactance is

Figure 16. Oscillator circuit for use with loop
antenna (T-172).

effectively introduced in the form of an off-
frequency external signal, it is necessary to
know the more complete behavior of the r-f
unit.

The two-terminal equivalent of an antenna
approaching a target is a fixed impedance or
admittance plus a rotating additional imped-

d This section may be considered as appendix ma-
terial to Section 3.1.2.

ance or admittance, provided the antenna does
not approach the target too closely. That is,
the incremental impedance or admittance pro-
duced by reflection of the antenna field changes
slowly in magnitude but undergoes a continu-
ous phase rotation (cf. Figure 1 of Chapter 2).
Under the normal operating condition wherein
the fixed portion of the antenna impedance is
real (resistive), the maximum and minimum
values of instantaneous detector voltage corre-
spond to phase angles of 0 and 180 degrees for
the incremental impedance. The complete com-
plex detector sensitivity to impedance changes
then reduces to a pure resistance sensitivity for
computing the magnitude of the audio-voltage
output.

The complete sensitivity could be expressed
in terms of either impedance or admittance.
The admittance evaluation is more convenient,
since combinations of resistance and reactance
can readily be placed across the fuze antenna
terminals in parallel with any antenna imped-
ance that may be present, whereas it is not
feasible experimentally to insert impedance ele-
ments in series with the antenna. Dealing with
admittances thus leads to an automatic ignor-
ing of the inherent shunt capacity from the
fuze cap to fuze ground.

If we write the antenna admittance as
A — C — jB, then the detector voltage will be
a function of both C and B. The term C is, of
course, the reciprocal of the parallel resistance
R in the formula S — R (dV/dR) . Therefore

V = V(C,B),
dV = dCdC + dB dB'

(17)

The physical condition that the incremental ad-
mittance is a rotating vector requires

so that

dC = | dC | cos 6,
dB = \dC\ sin 0,

Cw\ = cgeose + cgsine, (18)

indicating that dV is a sinusoidal voltage incre-
ment. We require its magnitude.

(19)

AMPLIFIER SYSTEMS

103

We so define these terms as to make the equa-
tion read

S =

y/ Sc2 + Sb2>

(20)

so that the complete sensitivity is given by the
quadrature addition of the conductance sensi-
tivity Sc and the susceptance sensitivity SB.

We immediately note that

1 dV
R d(l/R)

- RdV-
“ ~RdR ~

(21)

so that our former simplified definition is pre-
served when the susceptance (or reactance)
sensitivity vanishes.

Both quantities Sc and SB are readily meas-
ured in the laboratory as slopes of detector volt-
age versus values of antenna shunts.

There is one direct application of this formula
of interest to this section. It gives the solution
to the question of the effect of antenna circuit
detuning on sensitivity, a question of practical
significance in oscillator-diode fuzes.

Consider the fuze as a constant-current gen-
erator feeding the tuned diode-antenna circuit,
with internal admittance Ai — Ct — jB,. Then
the r-f voltage developed will be

E =

(22)

l(Ct + C) - j(Bi + B)Y
and the diode will yield a detector voltage pro-
portional to the magnitude of Elt

V = kI[(Ci + Cy + (Bi + B)2]~K (23)

When the fuze is properly tuned (maximum
detector voltage) we have

V = kI(C + Ci)~\

rdV -kIC
dC (C + C;)2

where V0 is the value of V when C — 0, i.e., the
idling voltage. This is our original sensitivity
formula, except for the trivial change of sign.
If the fuze is detuned,

V = kl [(C + Ci)2 + (B + Bi)2]~\

C % = - kIC(C + Ci) [(C + cy + {B +

_ -C(C + CQV» _ „

C kiy ~ Sd’ (25)

where SD represents the sensitivity when the

fuze is detuned. We already have for the tuned
sensitivity

(26)

in terms of the tuned voltage. Thus the ratio

Sd = C + Cj JP = JP

ST kl VT2 Vtv 1 ;

shows that the sensitivity falls off with detun-
ing as the cube of the voltage.

But if the voltage has been decreased by mak-
ing B -f Bt 0, then the response to variations
in B must be taken into account. We must com-
pute the complete sensitivity

Now

s = \4sc2 + sB 2.

Sb = c % = ~kIC(B + s*)-

(28)

[(c + cy + (b + Bym

C{B + Bi) V 3

(kiy

(29)

and

s =VsB 2 + sc 2 = VSb2 + sy

— jjepy \/(b + By + (c + cy

to be compared with

PC
kl ’

(30)

yielding

« _ Vr2C
6 T kl ’

S V2
ST ~ V t2’

(31)

(32)

so that actually detuning drops the sensitivity
only as the square of the voltage and not as the
cube. This is important in setting specification
limits on the accuracy of tuning in production.

3-2 AMPLIFIER SYSTEMSe

321 General Requirements

The amplifier receives the signals from the
r-f section of the fuze and is required so to
modify them that the desired signal will operate

e This section was prepared by Bertrand J. Miller,
Ordnance Development Division of the National Bureau
of Standards.

SECRET

104

ELECTRONIC CONTROL SYSTEMS

the thyratron at the appropriate time and place.
Since the amplifier input signal consists of sev-
eral components (desired signal due to reflec-
tion from target, noise due to tube and circuit
vibrations, hum due to a-c filament operation
and imperfect filtering of B supply voltage,
etc.), these modifications consist of the follow-
ing changes.

Amplification of the Desired Signal. In most
applications, the signal generated by the r-f
section is small compared to variations in strik-
ing voltage (critical bias) of the thyratrons,
due to tolerances in manufacture, variations in
supply voltages, temperature and other operat-
ing conditions. The amount of amplification re-
quired is different for fuzes for different appli-
cations and different for different trajectories
encountered with the same application. Thus
fuzes for different purposes require different
amplifiers. The variation with trajectory usually
imposes a requirement on the shaping of a cer-
tain sector of the gain-frequency curve of the
amplifier, since the different trajectories are
generally characterized by different signal fre-
quencies at the desired point of operation. The
maximum voltage gain required in the fuzes
developed by Division 4 has usually been of the
order of 150 times; the frequency region con-
taining the desired signals has been between
50 and 350 c.

Attenuation of TJndesired Signals. The most
prominent signals, aside from those due to
presence of the target, are microphonic noise
and hum due to a-c operation. The latter, of
course, consists of an approximately sinusoidal
signal, of fundamental frequency varying from
700 c up to, in some cases, several thousand
cycles. The amplitude is generally of the order
of the filament supply voltage, that is, 1 to
IV2 v. After sufficient refinement in oscillator
tube design, the microphonic noise was also
restricted to high frequencies, generally above
2,000 c. Even after all refinements of tube con-
struction and selection processes developed to
date, considerable noise in the high-frequency
region can be expected. Under the severe vi-
bration conditions encountered sharp spikes
of the same order as the hum voltages can still
be expected from the most carefully chosen
tubes.

The preceding considerations impose two ad-
ditional design conditions on the amplifier. In
order to reject these undesired signals, which
are of the order of volts, and function on the
desired signal of the order of hundredths of a
volt, the amplifier is required to have a sharp
high-frequency cutoff. In addition, the ampli-
fier is required to be linear up to large input
voltages at high frequencies to avoid genera-
tion of voltages in the pass band by rectification
of noise and hum envelope variations or by
generating difference-frequency terms from two
nearly equal noise or hum voltages or their
harmonics. (The presence of two mechanically
independent filaments in the triode oscillator
made this last circumstance seem especially
likely. Laboratory vibration tests showed that
shock excitation of the filament resonances is
very common, and that the two resonant fre-
quencies generally differ very slightly, by an
amount frequently in the signal-frequency re-
gion.) Of course, any serious overload at the
always present, hum voltage frequency and
magnitude would keep the amplifier perma-
nently paralyzed and prevent amplification of
signal frequencies.

Finally then, all these requirements are to be
met in an amplifier which is compact, not criti-
cal either to supply voltages or to variations in
component values due to manufacturing toler-
ances, insensitive to very wide ambient tem-
perature and humidity conditions, both during
the short time of use, and for long periods of
storage, rugged enough not to generate noises
of its own, under conditions of severe vibration,
and in some cases capable of withstanding ac-
celerations up to 12,000#.

3'2'2 Selection of Amplifier Characteristics

Three general types of amplifier characteris-
tics are required: one for the longitudinally
excited fuze for use against airborne targets,
one for the longitudinally excited fuze for use
against ground, and one for the transversely
excited fuze for use against ground. The ruling
factors and the resultant characteristics are
quite different, so the three types will be dis-
cussed separately.

AMPLIFIER SYSTEMS

105

Antiaircraft Target

The central problem in the case of an air-
borne target is the design of a fuze, usable on
a variety of rockets of different physical dimen-
sions, to produce a burst when the rocket passes
approximately abeam of its target (see Section
1.3). Relative velocities between target and
missile of 700 to 1,900 fps can be expected.
Function near the ideal burst surface is desired
out to passage distances of 70 ft or more.

The signal input to the amplifier under these
conditions has been discussed in Sections 2.11.2
and 2.11.3.

The important characteristics are decrease of
signal frequency from a maximum of 2 Vfk
(frequently called the head-on doppler fre-
quency) to zero and an increase of signal ampli-
tude ; most of both changes take place in a lim-
ited region near the target. Thus a peaked
amplifier with maximum gain somewhat below
the head-on doppler frequency would tend to
localize bursts in the appropriate region. Too
sharp an amplifier cannot be used, since calcu-
lations show that, in the region where burst is
desired, the signal frequency is changing rap-
idly (as high as 50 per cent change in frequency
per cycle). The breadth of the amplifier re-
quired to realize much gain on such a signal
would presumably make the precise location of
the peak frequency less critical. The best loca-
tion for the peak and the gain required were
determined empirically.

The empirical studies consisted of field tests
and of laboratory tests with the “drum genera-
tor” on the audio-signal simulator discussed in
Section 2.12. As a result of such tests, the fol-
lowing factors were established :

1. For an r-f sensitivity of approximately 15
v and a fuze directivity pattern approximately
like that of a half-wave dipole and a carrier
frequency in the vicinity of Brown reference,
the required amplifier sensitivity should be such
that about 30-rms-mv input signal (at fre-
quency of peak gain) will fire the thyratron.

2. The peak frequency can be located almost
anywhere below two-thirds of the head-on dop-
pler frequency with reasonably good burst
placement. For function nearly abeam of the
target rather than earlier on the trajectory
(see footnote b of Chapter 1) the amplifier peak

frequency equal to about one-half the head-on
doppler would be optimum. One-half is a nom-
inal value, since the head-on doppler frequency
obviously varies with rocket velocity, target
velocity, launching plane velocity, and relative
orientations of these. The normal figure refers
to the case of medium rocket velocity, equal at-
tacking plane, and target plane velocities, and
an overtaking aspect for the rocket. Any other
orientation of target and rocket velocities would
give a higher head-on doppler frequency, and a
smaller value of the ratio of peak frequency
to head-on doppler. A smaller value of this ratio
would give bursts closer to the ideal surface in
the overtaking aspect, but would probably be
too low for the nose attack and other high
relative-velocity aspects. Not much experimen-
tal information is available on this point, since
most testing was done with a stationary target,
and it was not possible to mock-up the head-on
aspect by adding twice a combat-plane velocity
to the normal rocket speed.1

A cutoff on the high-frequency side at a rate
which reduces the gain by a factor of 10 at one
and one-half to two times peak frequency and
continues on down at a slightly smaller slope
was found to be fast enough, reducing the gain
to approximately unity at power-supply fre-
quencies. Adjustment of the low-frequency side
of the gain curve is ordinarily made to give a
half-gain width of approximately half the peak
frequency. Drum generator studies (see Sec-
tion 2.12) show that at this width the gain of
the amplifier to the type of signal actually en-
countered in use is nearly the same as the
steady-state gain for a sine wave of the same
instantaneous frequency for frequencies in the
vicinity of peak, and that objectionable delays
are not encountered.

The manner in which the circuit problems
incident on realization of an amplifier having
the above characteristics were solved will be
discussed in Section 3.2.3. Time may be taken
here to point out, however, that a different
solution is possible, as pointed out in an early
NDRC report.1 This solution involves using
only the decreasing frequency characteristic of
the signal. The signal wave form is amplified
and clipped into a square wave; the duration
of each cycle is then measured. Firing of the

SECRET

106

ELECTRONIC CONTROL SYSTEMS

thyratron is accomplished when a long enough
half-cycle occurs. This attack was not pursued
at that time inasmuch as it required the use
of two tubes, whereas an amplifier could be
designed to give reasonably good burst place-
ment with a single pentode. Where space is
available, however, the alternative solution may
have other advantages which warrant further
investigation of this approach.

Some study has also been made of amplifier
requirements for fuzes suitable for air-to-air
bombing. Here one deals with low relative
velocities, and, in addition, a different orienta-
tion of relative velocities in the most important
tactical case. For the rocket case, with emphasis
on the overtaking aspect, the rocket velocity is
along the rocket axis in a coordinate system
fixed in the target. This is a fortunate situation
in one respect, since this state of affairs can be
simulated in field tests with a stationary target.
This state of affairs does not exist in the case
of bombing a formation from above. The result
of the difference is a slower rate of increase of
signal from the r-f section, and, in general, a
displacement of the point of maximum signal
away from the point on the trajectory of closest
approach. Details of the computations are re-
ferred to in the bibliography;113 no experi-
ments were carried out by Division 4.

Ground Approach, Longitudinal Excitation

The problem here is the development of a
fuze which will give substantially uniform burst
heights on a variety of bombs, dropped from
different altitudes and at different airplane
speeds; the same fuze to be useful both for
level-flight release and dive bombing if possible.
In order to show the degree to which it is pos-
sible to harmonize these requirements by ap-
propriate amplifier design, the requirements
for one fuze of this type will be presented in
detail.

We choose for the purpose of this illustration,
the requirements for a fuze operating at 75 me,
with use on the M-30 (100-lb GP) and the M-81
(265-lb fragmentation) bombs being contem-
plated. This fuze is also usable on the M-66
(2,000-lb GP), and with somewhat less effec-
tiveness, on the M-64 (500-lb GP) ; but here we
will consider only the first two bombs.

Because of the varying degree of mismatch
between oscillator impedance and radiation
load, the fuze r-f section will not have the same
r-f sensitivity on the two bombs; we take as
representative values a sensitivity of 11 v on
the M-30 and 14 v on the M-81. Further, al-
though the variation is slight, the directivity
patterns are somewhat different, due to the dif-
ferences in effective electrical length; the por-
tions of the pattern of tactical interest are
shown in Figures 17 and 18.

Figure 17. Directivity pattern for M-30 bomb
at Brown frequency, longitudinal excitation.
Curve shows detail for small angles off nose of
missile.

In addition, the ballistic coefficients of the
two bombs differ, so that similar release condi-
tions result in slightly different terminal con-
ditions (velocity and angle of approach to the
ground) .

Function heights of the order of 2 to 7 wave-
lengths will be considered; for some of these,
near the nulls of the directivity pattern, it is
necessary to consider the contribution of the
induction field (inverse R2 field, see Section
2.10). For this reason, function heights for
nearly vertical approaches do not vary directly
with amplifier gain.

Computation will be made of the amplifier
gain necessary to give a function height of 50 ft
over a surface with reflection coefficient equal
to 0.7 for various combinations of release alti-
tude and plane speed, both for level flight and
dive bombing. The computations are made on

AMPLIFIER SYSTEMS

107

the basis of the theory developed in Sections 2.9
and 2.10.101> 140

Curves of voltage gain versus frequency re-
quired for these various conditions, based on

Figure 18. Directivity pattern for M-81 bomb,
Brown frequency, longitudinal excitation. Curve
shows detail for small angles off nose of missile.

an assumed holding bias of 5.3 v on the thyra-
tron, are shown in Figure 19. The frequency
ranges shown on each curve correspond to re-
lease altitudes from 2,000 to 20,000 ft for the
level-flight cases, and 1,000 to 10,000 ft in the
dive-bombing cases, which is assumed to con-
tain all the range of interest. Outside these
ranges, the gain should be low.

Examination of Figure 19 shows that the
requirements for different release conditions
are conflicting, so that a compromise is neces-
sary. In making such a compromise, the follow-
ing considerations are general :

1. The high-frequency end of each curve cor-
responds to a steeper angle of approach than
the low-frequency end. At steeper angles the
induction (inverse R2) field is more important.
Consequently, the height of function varies
more nearly with the square root of gain at
high frequencies than at low. This gives greater
freedom of design in this region of the spec-
trum.

2. If an oscillator-diode type of fuze is con-
templated, the tuning problem must be consid-
ered. In the case of the fuze under discussion,
all production models were oscillator-diode.
These were tuned on the M-30, so that the full
11-v sensitivity assumed was probably very

closely realized on that bomb. Because of a
slight difference in reactance of the two ve-
hicles at the feedpoint, however, the fuze was
somewhat detuned on the M-81, resulting in a
reduction of the average sensitivity by about
5 per cent.

3. Very high gains should be avoided as far
as possible without loss of effectiveness, since
high gain obviously increases the probability of
malfunction.

One compromise actually used is also shown
on Figure 19, the curve being an average curve
for production units (Philco T-91, type 20 am-
plifier). This fuze was designed in response to
a request to give special weight to low-altitude

Figure 19. Amplifier gain curves required for
longitudinally excited fuzes for different bombs
in different release conditions.

The labeled curves represent conditions as follows: A, M-30
bomb released at 200 mph in level flight; B, M-30 bomb
released at 300 mph at level flight; C, M-81 bomb released
at 200 mph in level flight; D, M-81 bomb released at 300
mph in level flight; E, M-81 bomb released at 400 mph in
60-degree dive; F, M-81 bomb released at 300 mph in 30-
degree dive. Full curve represents a compromise gain-
frequency characteristic for typical unit.

and diving releases. The corresponding func-
tion heights are shown in Figure 20, for the
level flight cases and in Figure 21, for the dive-
bombing cases.

108

ELECTRONIC CONTROL SYSTEMS

Since the input signal in this application does
not change in frequency or amplitude rapidly
in the region where function is desired, steady-
state calculations are adequate. However, for
the purpose of determining amplifier delays,
more precise calculations were made, making
use of Borel’s theorem. According to the theo-
rem, the response of any linear network can be
computed for any form of input if one has
either the network response to a unit step H (t)
or a sharp spike of unit impulse H'(t ).54 Fig-
ure 22 shows H(t) and H'(t) for a typical am-
plifier, and Figure 28 the delays computed. The
computed delays are less than 5 ft for all tac-
tical situations for the amplifier, and are not
longer for the other amplifiers employed.

For the fuzes at other frequencies, the re-
quirements are very similar and will not be
detailed here. The similarity extends even to
maximum gain required, so that the changes
consist primarily in frequency shifts, signal

O 4000 8000 12000 16000 20000

RELEASE ALTITUDE (FEET)

Figure 20. Function heights computed from
gain curve of Figure 19 for level flight release.
The various curves represent conditions as
follows: A, M-30 bomb released at 200 mph; B,
M-30 bomb released at 300 mph; C, M-81 bomb
released at 200 mph; and D, M-81 bomb released
at 300 mph.

discussion. Here, lower function heights are
desired, of the order of 10 to 15 ft. The projec-
tiles are much smaller than bombs and short
compared to the carrier wavelengths proposed.

Figure 21. Function heights computed from
gain curve of Figure 19 for dive releases. For
M-81 bombs, curve E is for a 60-degree dive at
400 mph, and curve F is for a 30-degree dive at
300 mph.

As a consequence the radiation resistance is
high and varies rapidly with frequency. The
directivity pattern, however, is nearly inde-
pendent of frequency for carrier frequencies
below 135 me. Thus for low-carrier frequency,
one expects low r-f sensitivity, but this is bal-
anced by a larger scale factor (wavelength, A),
and more important induction field contribution.
(The latter is of great importance because of the
low function height desired.) As a consequence
of the interaction of these factors, it develops
that an amplifier gain curve can be drawn
which is optimum not just for one carrier fre-
quency but for any carrier between 70 and
130 me. It was also found possible to realize this
gain-frequency curve (Figure 24) econom-
ically.130

frequencies being proportional to carrier fre-
quency for bombs with similar ballistics.

The fuze for the mortar projectile using lon-
gitudinal excitation merits a brief separate

Ground Approach, Transverse Excitation

The difference between the amplifier required
in the case of a transversely excited fuze and
that of the longitudinally excited fuze arises

SECRET

AMPLIFIER SYSTEMS

109

from the difference in the directivity patterns.
In the longitudinal case, one has axial sym-
metry about the axis of the bomb. The direc-
tivity pattern is a minimum for vertical ap-

Figure 22. Response of typical amplifier to a
unit step function, H (t) , and a short pulse of
unit impulse, H' (£).

proach, and increases rapidly as the angle of
approach increases (angle between trajectory
and vertical) for any orientation of the bomb
about its axis.

In the case of the transversely excited fuze,
the directivity pattern is maximum for vertical
approach. For angles of approach other than
zero, the value of the directivity pattern de-
pends somewhat on the orientation. For most
tactical situations, however, the signal strength
is nearly independent of release altitude and
hence of signal frequency. A relatively flat gain
curve is therefore required. For the carrier fre-
quency used (about 150 me), and the bombs
employed, the useful tactical range of altitudes
gives a frequency range from 165 to 330 c.

The value of the maximum gain is determined
by the r-f sensitivity at the operating load and

by the directivity pattern. Because of the high
radiation resistance, and consequent poor match
to oscillator impedance, sensitivities are lower
than those encountered on most bombs with the
longitudinal fuzes. The antenna gains were ap-
proximately the same. However, the tactical
range of approach angles centered near the
maximum of the directivity pattern in the
transverse types, instead of near the minimum,
as was the case with the longitudinal types.
The net effect of all these factors is a require-
ment for somewhat less gain (for the same
height) for the transverse fuzes. Production
fuzes, however, were in fact built with approx-
imately the same maximum gain as the longi-

Figure 23. Amplifier input and output signals
for various heights above ground, assuming ap-
proach of 30 degrees with the vertical and
vertical velocity of 790 fps. Assumed amplifier
peaked frequency is 120 c. Carrier frequency:
Brown. Solid curve represents input signal multi-
plied by steady state gain; dashed line repre-
sents output signal (inverted).

tudinal fuzes and consequently gave somewhat
greater heights of function.

This amplifier is required to have a sharper
high-frequency cutoff than the amplifier for

SECRET

110

ELECTRONIC CONTROL SYSTEMS

the longitudinal fuzes by virtue of the higher
carrier frequency employed. This necessitated
maintaining a high gain at 330 c, whereas the

Figure 24. Gain-frequency characteristic curve
required for trench mortar shell fuze, assuming
carrier frequency compensation.

highest frequency encountered with longitu-
dinal types was 225 c. Since power-supply fre-
quencies at arming may be as low as 700 c, a
very rapid cutoff is required.

3 2 3 Methods of Securing Required Gain
and Shaping; Typical Circuits

Axial Antenna Fuzes

All amplifiers used in modern fuzes for non-
rotating missiles are lineal descendants of the
amplifier developed for the MC-382, the elec-
tronic control part of the T-5 or T-6 fuzes. Here
the shaping was accomplished by a feedback
network similar to that employed in RC oscil-
lators, using only resistors and condensers. With
a network employing three series condenser-
shunt resistor sections between grid and plate
of single tube amplifiers, the following relations
are noted :

1. At low frequencies, the feedback ampli-
tude is very small and phase shift is nearly
270 degrees.

2. At higher frequencies, feedback amplitude
is larger and phase shift through network is in
the vicinity of 180 degrees; this constitutes
regenerative feedback.

3. At still higher frequencies, feedback am-
plitude is still larger and phase shift approaches
zero, i.e., the amplifier becomes highly degen-
erate.

This principle is attractive because it gives
promise of providing a sharp high-frequency
cutoff, together with large voltage-handling ca-
pacity in the high-frequency region, with the
consequent freedom from cross modulation of
large noise voltages. Consequently, the circuit
was developed and used in the form shown in
Figure 25, employing pentodes developed from
hearing-aid tubes.

One peculiarity of this circuit is perhaps
sufficiently basic to warrant discussion here.
The resistances Rg, R7, and the parallel com-
bination of Rg and the source impedance, con-
stitute a voltage divider which controls the
amount of feedback. Since RG is of the order of
megohms, and the impedance of the r-f section
as an audio generator is of the order of tens
of kilo ohms, it is evident that the amplifier
characteristics depend markedly on the imped-
ance of the source. Thus all amplifiers of this
family required properly designed test circuits
(see Chapter 7) which simulated the impedance

Figure 25. Schematic circuit of feedback ampli-
fier employed in T-5, T-6 fuzes.

of the sources for which the amplifiers were
designed. Similarly, some restrictions were im-
posed on source design, since the amplifier pre-
sented to the source an impedance varied radi-
cally with frequency, even becoming negative
in certain cases.

The values used in this circuit will not be
cited as typical examples, despite the historic
interest of the circuit, for two reasons. First,
the circuit was the last designed for battery

AMPLIFIER SYSTEMS

111

operation ; the high-frequency cutoff was inade-
quate for generator use with raw alternating
current as a filament supply. Second, because
of the status of tube de\*elopment at the time,
many of the values represented compromises.
Tube development and circuit development were
proceeding simultaneously. At the time epoch
corresponding to Figure 25, two reasonably
satisfactory but different pentodes had been
developed by different laboratories. The di-
vergence in characteristics, however, was not
so large that using both in the same amplifier
was not feasible, by some judicious compromis-
ing on values.

A more typical amplifier, therefore, is shown
in Figure 26. This is a type furnishing a gain-
frequency curve of shape suitable either for
airborne targets or for ground approach, with
longitudinal excitation. It will be noted that
the feedback voltage is now divided by a ca-
pacity, rather than a resistance divider; at
frequencies above peak frequency, this provides
capacity loading on the input grid and improves
the high-frequency cutoff. Further high-fre-
quency attenuation is provided by the series

tion. This scheme was adopted because it proved
to be possible in this way to control peak gain
with only minor effects on the frequency at
which peak gain was realized.

A typical gain-frequency curve is reproduced
as Figure 27 ; also shown is the curve noted
when the pentode grid side of both feedback

FREQUENCY (CPS)

Figure 27. Gain-frequency characteristic and
flat gain characteristic given by circuit of Figure
26.

Figure 26. Later feedback amplifier circuit
employed in generator-powered fuzes. Feedback
circuit shown employs a feedback divider (Cu,
Ci6> and gain control (Cs) .

R-shunt C network in the thyratron grid cir-
cuit. A higher gain level was sought and was
obtained in part by increasing the feedback.
This additional gain necessitated the provision
of a feedback adjustment. A variable capacity C
was provided for this purpose. The feedback
network is designed to give too much feedback,
so that gains are too high; adjustment of C
introduces a controllable amount of degenera-

loops is disconnected. It will be noted that the
gain at peak frequency is multiplied by a
factor of about 2.5 by the feedback; this is
about the maximum amount of feedback usable
if too sharp gain curves and undue dependence
on variations in supply voltage are to be
avoided. The gain without feedback is of course
depressed by the various high-frequency atten-
uating networks, and by the loading of the
plate circuit by the feedback loop; the same
tube in a conventional RC-coupled amplifier
gives a gain of about 100 times with the supply
voltages indicated.

Within reasonable limits, the amplifier can
be redesigned to give the same shaping at other
peak frequencies by simply scaling the capaci-
ties; this practice was largely followed here,
with minor readjustment of values to commer-
cially available ones where necessary.

A somewhat different shaping was required
for mortar fuzes, because of the rather differ-
ent ballistic properties and conditions of use
for these projectiles ; the requisite gain curve is
shown in Figure 24.

SECRET

112

ELECTRONIC CONTROL SYSTEMS

Trans verse- Antenna Fuzes

The relatively high flat plateau, with steep
cutoff on the high side, required of amplifiers
for transverse-antenna fuzes suggests the use
of two peaks. This attack on the problem has

TO RF

Figure 28. Two-stage feedback amplifier for
use with transverse antenna fuze.

been pursued in various ways. Perhaps the most
obvious is the use of two stages similar to Fig-
ure 25 in cascade, with the feedback loops ad-
justed to different peak frequencies. Another
possibility is the use of two feedback loops in

Figure 29. Gain-frequency characteristic for
two-stage feedback amplifier of Figure 28.

the same stage. Here a network similar to that
of Figure 25 was employed for the higher
frequency peak, and a network with series re-
sistance and shunt capacity was used for the
lower frequency peak. (This order was essen-
tial since the degenerative feedback below peak

frequency for the series C network was atten-
uated, as was that above peak-frequency for
the shunt C network.) Finally, the conventional
feedback network to. provide the high peak,
combined with an LC network for the lower-
frequency peak, could be used.

All these approaches were investigated. Lim-
ited experience with the dual-feedback loop
indicated a relatively high supply-voltage sen-
sitivity; since other successful and economical
solutions of the problem were available, this
attack was not vigorously pressed. A typical
two-stage solution is shown in Figure 28. Be-
cause of requiring two tubes, this solution
found little use ; however, it obviously has

Figure 30. Schematic diagram for feedback
amplifier with tuned choke input (used in T-51).

greater flexibility and, as shown in Figure 29,
can provide extremely sharp cutoffs which
might be necessary in some applications.

The solution of the problem which found the
greatest practical use employed a series reso-
nant LC network in the pentode grid for the
low-frequency peak. The high-frequency peak
was obtained by feedback ; the two circuits com-
bined to give a very fast high-frequency cut-
off, necessary in this case by virtue of the prox-
imity between signal and generator frequencies.
(This same fact was responsible for the fila-
ment center tapping employed.) The feedback
circuit thus still supplies the high-frequency
grid degeneration which enables the amplifier
to handle large high-frequency signals. The
design of the feedback loop, as noted in Fig-
ure 30 is conventional; the same type of gain
control is used.

, SECRET

AMPLIFIER SYSTEMS

113

The low-frequency peak is supplied by the
grid choke and C3; the Q of this resonant cir-
cuit is controlled by a series resistor. It is of
interest to note that a low-impedance C bias
source must be provided for in order not to
broaden the resonance curve.

Figure 31. Gain-frequency characteristic for
choke input feedback amplifier.

The type of M-8 head employing plate cur-
rent rather than grid voltage variations was
used in one bomb fuze with transverse excita-
tion (T-82) and thus required a similar am-
plifier-gain curve. Figures 32 and 33 show a
typical circuit and gain curve. Here, trans-

Figure 32. Schematic diagram for feedback
amplifier with transformer input. (The switch
shown below C15 is a sensitivity control.)

former input to the amplifier is employed, the
low-frequency peak being supplied by resonat-
ing the transformer secondary ; the usual high-
frequency feedback network being employed.

Combination Amplifiers
The applicability of a fuze to various missiles

could be increased if it were possible to vary
the amplifier shaping in the field by a simple
adjustment. Thus, although the same general
shaping is required in a rocket fuze for ground
approach and for airborne targets, the required

Figure 33. Gain-frequency characteristics of
amplifier shown in Figure 32. High and low
curves correspond to switch open and closed.

peak frequency and gain are different. The
same remark is true if it is desired to use the
same type of fuze against airborne targets on
rockets and on bombs. By the introduction of
shorting switches, to cut additional feedback
sections in or out, good approximations to two
different ideal gain curves can be provided in

Figure 34. Combination amplifier for use in
air-to-air and air-to-ground applications. Switch
S allows transfer from one application to the
other.

the same amplifier. Typical circuit and gain
curves are shown in Figures 34 and 35. Since

SECRET

114

ELECTRONIC CONTROL SYSTEMS

operation of the switch consists in removing
or inserting a screw from the chassis, the oper-
ation is readily performed in the field immedi-
ately before fuzing the projectile when the ap-
plication is determined.

A similar adjustment was investigated on
some transverse-type fuzes to provide sensi-

Figure 35. Gain-frequency characteristics for
amplifier shown in Figure 34. Upper curve repre-
sents air-to-air case, the other, air-to-ground.

tivity control. Removal of a screw inserts a
voltage divider in the amplifier, whose effect is
to halve the gain and thus halve the expected
height of burst.

324 Properties of Pentodes

The electric properties of the pentodes used
in the amplifiers differed in some respects ac-
cording to the manufacturer of the pentode.
Average values are shown in the accompanying
table.

Raytheon

Sylvania

ge2

2.0

1.6

1.4

218

195

176

445

310

250

Rscreen grid

0.32

0.27

0.54

Input

impedance

megohms

yumhos

megohms

megohms
(measured
at 60 c)

The values cited above were not measured at

the same element voltages for the different
tubes, but at the operating voltages occurring
at those elements when the tube was operated
in a typical feedback amplifier, with 1.4-v A
supply, 140-v B supply, and — 1.8-v C bias.

Security requirements were imposed on the
mechanical properties in part. (See Section
3.1.5.) For the greater part of the time interval,
including the development and tactical use of
fuzes for nonrotating missiles, it was required
that all tubes used must fail on a 20,000# cen-
trifuge test. It proved possible to build tubes
which would pass 2,500# with reasonable as-
surance of failure at 20,000# ; accordingly, this
level of ruggedness was chosen for the great
majority of the tubes built for this program.

Figure 36. View of three pentodes used in
amplifiers for proximity fuzes. These are from
left to right: GE pentode, Raytheon pentode
NR-5, and Sylvania NS-5 pentode.

In the latter stages of development, some mis-
siles with launching accelerations near 10,000#
were encountered. Relaxation of security re-
quirements was secured and more rugged tubes
were made by simple mechanical changes.

In favorable contrast to the oscillator triode
situation, the allowable ruggedness level for the
pentode proved sufficient for suppression of
microphonics. Presumably because of the low
voltage level at the pentode grid, microphonic
audio amplifiers were exceedingly rare. In al-
most every case of a noisy unit, blocking of the
oscillator would remove all microphonic output.

Figure 36 showTs the external appearance of
pentodes of different manufacturers ; Figure 37
shows the internal construction of a typical
pentode.

AMPLIFIER SYSTEMS

115

For further details reference should be made
to the final reports of the tube manufactur-
ers.201’ 202

3.2.5 Adjustment and Testing

With components of commercial tolerances,
it was found possible to build the amplifiers
with only one adjustable component; the in-
verse feedback condenser, shown, for example,
as C8 in Figure 26. This condenser controlled

Figure 37. View of NS-5 pentode with envelope
removed, showing various components.

the peak gain primarily, with only second-order
effects on the peak frequency. In some of the
double-peaked amplifiers, an additional adjust-
ment was provided in the form of a resistor
controlling the Q of the series resonant circuit
responsible for the low-frequency peak.

Testing is described in greater detail in
Chapter 7. Two properties of the signal injec-
tion circuit were critical: the impedance and
the hum level. The impedance of the synthetic
signal source had to be the same as the actual
source, since, as was pointed out earlier, this
impedance is a portion of the feedback voltage
dividing network. In addition, hum, or a signal

of power-supply frequency, had to be injected
along with the desired signal to simulate actual
operating conditions. This was because milli-
volts required to fire the thyratron were meas-
ured rather than voltage gain, since the former
were of more direct applicability. The effective
critical bias on the thyratron, however, de-
pended on the amount and phase of hum voltage
passed through the amplifier. Special signal in-
jection circuits were consequently designed for
each amplifier.

Where gain was measured, the output volt-
age was defined as that appearing on the thyra-
tron grid. In virtue of the feedback applied to
the amplifier and also of the high-frequency
attenuating network preceding the thyratron
grid, the impedance level at this point was very
high. This necessitated the use of very high-
impedance voltmeters for the measurement of
output voltage.

3 2 6 Response to Spurious Signals

Spurious signals are here defined as any sig-
nals due to causes other than motion with re-
spect to a reflector. Since these latter are
always expected in a relatively restricted re-
gion of the frequency spectrum, the first pre-
caution is evidently to keep amplifier gain low
outside this region. Certain transients however
are of sufficient magnitude to require separate
consideration.

The rocket fuzes for antiaircraft use gener-
ally were required to be ready for operation in
a period ranging from % sec to slightly more
than 1 sec. During this period, the d-c level at
the amplifier input changes from 0 to approxi-
mately — 40 v; filament voltage is applied sud-
denly to all tubes and plate voltage to all, al-
though not simultaneously; the application of
plate voltage to the thyratron is delayed. Firing
during this cycle is prevented by maintaining
the following sequence.

1. Plate and filament voltage to oscillator,
diode (if any), pentode, and filament voltage
for the thyratron are applied. Since the pentode
filament is not yet emitting, its plate assumes
the potential of the supply voltage and a posi-
tive pulse appears on the thyratron grid. Since

SECRET

116

ELECTRONIC CONTROL SYSTEMS

the thyratron filament is also cold and its plate
circuit is open, firing does not occur.

2. The properties of the pentode and oscilla-
tor (and diode, if any) filaments, and the volt-
ages supplied to them, are so adjusted that the
oscillator (or oscillator and diode) warm up
before the pentode. Thus when the pentode
begins to warm up, the tube is at first cut off by
the negative surge on its input. Pentode warm-
up is substantially complete before this nega-
tive charge leaks off via the grid leak, resulting
in a smooth drop of the pentode plate to its
operating point. Thus, during the time the
thyratron is warming up, a negative signal is
appearing on its grid.

3. The only possible transients due to sud-
den application of thyratron plate voltage are
the signals due to thyratron plate-grid capaci-
tance, and any signals due to switch action in
leakage r-f fields from the oscillator. The first
is suppressed by the large capacity from thy-
ratron grid to ground. The second is suppressed
by associating chokes and capacities with the
switch in the appropriate fashion. Since leak-
age fields are small, switch-oscillator coupling
is not strong and circuit design is not critical.
(See Section 3.3.)

A somewhat different problem occurs in the
case of mortar fuzes. These are sometimes fired
at very high angles, so that velocities are low
near the peak of the trajectory. At the conse-
quent low generator speeds and supply volt-
ages, the oscillator plate current will be re-
duced; since plate current is a large factor in
determining C bias, the thyratron bias will be
reduced. Additionally, the supply-voltage rise
with increasing speed on the downward leg of
the trajectory will be very rapid and may give
rise to transients within the amplifier itself
resulting in positive signals on the thyratron
grid.

Since thyratron plate voltage is also low at
low speeds, some reduction in C bias can be
tolerated. A shunt load on the B supply can be
used which will draw enough current through
the bias resistor to hold the thyratron at low-
plate voltages.

The only transient on the downward leg of
the trajectory which gave positive thyratron-
grid signals was found to be associated with the

pentode screen-grid circuit. The problem was
solved by supplying the screen from a voltage
divider, in place of a simple series dropping
resistor. Because of the differing screen im-
pedances of tubes of different make, a different
divider was used for each tube manufac-
turer.

As in the rocket fuze, any transients due to
the oscillator dropping out of oscillation and
starting up again were handled by the shorter
warm-up time of the oscillator filament, com-
pared to pentode and thyratron filaments.
Where necessary, this difference was accentu-
ated by series resistors in pentode and thyra-
tron filament circuits.

It is of interest to note the rapidity of warm-
up achieved with the tubes at hand ; the
MC-382, for example, was completely stabilized
and ready for arming (application of thyratron
plate voltage) in less than 0.25 sec after the
application of oscillator and amplifier supply
voltage.

3,2 7 Tolerance of Components and
Variation in Performance

Exhaustive studies of the effects of varia-
tions in the component values were conducted.
In general, the conclusion reached was that
satisfactory restriction of performance varia-
tions could be achieved in the single-peak am-
plifiers by the use of unselected components of
10 per cent tolerance. (In the double peak am-
plifiers, 5 per cent components were used in the
feedback network.) By the use of sorting (pair-
ing high capacities and low resistors, or vice
versa) 20 per cent components could be used in
most places.

Except for a few small resistors (as in fila-
ment circuits), carbon resistors and paper con-
densers were employed. The temperature co-
efficients are opposite but that of the resistors
dominates, so that the amplifier has a negative
gain-temperature coefficient. The value depends
largely on the form of the gain-control con-
denser, ranging from —0.7 per cent per degree
centigrade with one form (twisted pair of
wires) to —0.2 per cent per degree centigrade
with another (ceramic condenser).

SECRET

THE DETONATOR CIRCUIT

117

The supply voltage sensitivity was not so
large as might be expected from a regenerative
circuit; percentage gain changes were approxi-
mately equal to percentage supply-voltage
changes.

Because of the many high-impedance points
m the amplifiers, good protection against mois-
ture was required, including “built-in” mois-
ture as well as any encountered in subsequent
exposures to humid atmospheres. The imped-
ance between pentode grid and plate was par-
ticularly high ; this necessitated specifying bet-
ter than common practice in tube washing to
eliminate any conducting salts or acids on the
tube press. The assembled amplifier was given
a dip in hot wax to drive out as much moisture
as possible and seal it out, after which the as-
sembly was potted. Tung oil and Glidden pot-
ting compounds were used. (See Section 4.7.)

Figure 38 shows the variations encountered
from all causes in one type of amplifier, except
supply voltages, which were standardized. Tem-
perature, of course, was not standardized, but
the range of variation was limited.

3 3 THE DETONATOR CIRCUITf
General Requirements

The purpose of any fuze is to initiate the
high-velocity shock wave needed to set off the
explosive charge. In the proximity fuze, an
electric detonator is used to link together the
operating parts of the fuze and the powder
train which sets off the high explosive. The de-
tonator assembly must meet the stringent
safety requirements of the Armed Forces, and,
for proper functioning, the detonator imposes
even more stringent requirements on the elec-
tric firing network. The detonator is fired by
the discharge through its bridge wire of the
energy stored on a large capacitor. A screen-
grid thyratron is used as the electronic switch
to discharge the capacitor through the detona-

f This section was prepared by Charles Ravitsky of
the Ordnance Development Division of the National
Bureau of Standards. Mahlon F. Peck of the same
organization prepared Section 3.3.4 on the properties of
the thyratron. Major bibliographical references for this
section are 6, 8, 25, 57, and 58.

tor. The thyratron holding bias is set so that
the tube will fire when the fuze receives a sig-
nal larger than the predetermined threshold for
functioning. In order that the fuze will not
function prematurely, both electric and me-
chanical methods are used to make the fuze en-
tirely inoperable before arming. Before elec-
tric arming occurs the detonator bridge wire is
not connected to the electric circuit, so that no
current can flow through it, regardless of what
happens to the rest of the fuze. Further, in the
generator-powered fuzes no power is available

\\\

/ 4

/>
/ /

\V\

/)
/ /

\\\

20 30 40 50 60 70 80 90 100 200

FREQUENCY (CPS)

Figure 38. Variations of gain encountered in
feedback amplifier. Voltages were held constant
and gain curves translated to common peak fre-
quency. Dotted curves represent limits within
which are included characteristics of approxi-
mately % of the amplifiers.

until the fuze is traveling through the air at a
speed greater than 80 mph (see Section 3.4).
As an additional protection, before mechanical
arming, a ^4-in. thick brass plate is located be-
tween the detonator and the tetryl powder train
(see Chapter 4). Thus, even if the detonator
explodes, the resultant shock wave, after pass-
ing through the brass plate, will not be large
enough to set off the tetryl.

It is obvious that the detonator itself must
be so located that the shock wave due to its de-
tonation will set off the tetryl booster charge.
This limitation permits either the end or a side
of the metal cup to face the tetryl. It is not so

SECRET

118

ELECTRONIC CONTROL SYSTEMS

obvious, however, that the electric components
of the detonator circuit must be so arranged
that stray coupling to the other electric net-
works is minimized. It was found that inade-
quate shielding of the detonator circuit some-
times caused early functioning of the fuze due
to transients set up at the time of arming. In
order to eliminate the detrimental effect of
transients when the detonator was connected
to the circuit at arming, r-f chokes were put
in series with the detonator, and a condenser
was used to by-pass it at radio frequencies
(compare with Section 3.1.5).

The principal components of the detonator
circuit are the electric detonator, the condenser,
and the thyratron (Figure 39). In addition, the
impedance of the filament power supply is in
series with one leg of the thyratron filament in
some of the fuzes, and half the impedance is in
series with each leg of the thyratron filament in

Figure 39. Elementary detonator circuit.

other fuzes. Also, in some of the fuzes there is
an r-f choke172 in series with each of the det-
onator leads and a small condenser in parallel
with the detonator. These additional compo-
nents have only a minor effect on the proper
operation of the detonator circuit and were
added to reduce transients at arming. A small
resistor of 3 or 5 ohms is included in series
with the ungrounded end of the thyratron fila-
ment in order to decrease the filament current
to the value for which the filament was de-
signed.

This section of the report will deal with the

major components of the detonator circuit. As
the other components do not appreciably influ-
ence the action of the circuit, they can be dis-
posed of here in only a few words. The filament
resistance is a commercial 1,4 -w resistor. The
r-f choke in series with the detonator is the
same as those used in the oscillator section of
the fuze. It consists of 78 turns of No. 38 enam-
eled wire, wound on a core the size of a *4-w
resistor. The shunting condenser used in the
T-82 is a 500 p|if Ceramicon condenser. The
impedance of the filament power supply is dis-
cussed in Section 3.4 on the fuze power supply,
which covers both batteries and generators.

The Detonator

The detonator, as the connection between the
fuze and the high explosive, is a critical part of
the fuze. In order for manufacture of the fuze
to be possible, the characteristics of the various
fuze elements must be such that they can meet
the requirements of the other components. The
initial requirement is imposed by the high ex-
plosive [HE] used in the missile. In order to set
off the HE the Army Ordnance Department re-
quired212 the use of a tetryl (trinitrophenyl-
methylnitramine) booster. To assure a high-
order detonation the Ordnance Department
specified that the tetryl powder train should
ample safety factor. The inner diameter of the
HE, and a length of 0.75 in. will allow an
ample safety factor. The inner diameter of the
tetryl cup may be as little as 0.375 in. with no
apparent diminution in the velocity of the
shock wave. With the tetryl powder train speci-
fied by the Army, the problem arose of procur-
ing an electric detonator which could initiate
a high-order detonation in the tetryl, and which
would not impose too severe requirements on
the electric components.

Many squibs, detonators, blasting caps, and
semicaps which were commercially available
were tried, as well as an experimental high-
impedance high-voltage detonator. For the ini-
tial fuze testing, the ND-5, a fast violent but
poor flame-throwing squib made by the Hercules
Powder Company, was used. This company
then developed the ND-24 for use in these fuzes,

THE DETONATOR CIRCUIT

119

as an improvement over the ND-5, Although it
functioned satisfactorily in the field tests, in
which it was used to set off a potassium per-
manganate or a black powder spotting charge
(cf. Chapter 8), it was not powerful enough
to insure detonation of the tetryl booster. It
therefore had to be abandoned. Hercules then
developed a satisfactory detonator, which was
known successively as the BS-4 or BS-5; the
Detonator, Electric, T-3; the Detonator, Elec-
tric, M-2; and finally the Detonator, Electric,
M-36. The latter three designations are official
Army Ordnance nomenclature and indicate ac-
ceptance for use, first as an experimental item
and then as a standard Army production item.

The M-2 electric detonator212 (see Figure
40) itself consists of a three-element powder
train. The bridge (heater) wire is embedded in
about 0.2 g of mercury fulminate, which is fol-
lowed by a primer charge of 0.13 ± 0.02 g of
lead azide, which in turn detonates the base
charge of 0.14 ± 0.02 g of pentaerythrite tetra-

Figure 40. M-2 detonator used in radio prox-

imity fuzes. This detonator was also designated
BS-4 and M-36.

nitrate [PETN]. This construction permits a
very compact assembly for the explosive pow-
der train leading to the tetryl booster charge.
The bridge wire, when heated electrically, sup-
plies the energy to ignite the mercury fulmi-
nate and thus initiates the explosion. The
heater wire is made of Nichrome and is ini-
tially 0.09 ± 0.01 in. long and is 0.00050 ±
0.00005 in. in diameter. In connecting the Ni-
chrome wire to the copper lead wires, solder is
dropped over the ends, so that only about half
its length is effective in heating the mercury
fulminate. The specified resistance for the deto-
nator is 9 ± 3 ohms, allowing a 100 per cent
variation between the minimum and maximum
resistances and therefore between the mini-
mum and maximum effective lengths.

Time Lags. As the fuzes are used on projec-
tiles traveling at high speeds, any time lag be-

tween the operation of the fuze and the explo-
sion of the HE must be quite small. The delay
inherent in the detonator may be broken down
into three components : the time it takes for the
incident electric energy to heat the Nichrome
wire, the transfer of enough heat energy to
ignite the mercury fulminate, and the explosion
of the PETN base charge, which detonates the
tetryl booster.

In operation, enough heat is generated in the
bridge wire to melt it by the dissipation of over
1 millijoule of energy in it within 1 msec.57
Nichrome melts at 1350 C, so that the grains of
mercury fulminate adjacent to the bridge wire
become immersed in a metallic bath at that
temperature. The time between the liquefaction
of the bridge wire and the explosion of the
PETN is less than 0.2 msec. The Nichrome wire
heats about 4 micrograms of mercury fulminate
to its ignition temperature to start the explo-
sion, which is thus initiated within a cylindrical
layer about 0.00045 in. thick around the
0.00025-in. radius Nichrome wire.57 The explo-
sive wave travels through the tetryl booster
after its initiation by the PETN, at over 7,000
m per sec, so that the time lag in the booster is
less than 5 psec. The time delay in the explo-
sive elements is thus quite small. The overall
time delay in the detonator is, however, influ-
enced by the rate at which electric energy is
dissipated in the Nichrome heater wire. If this
rate is less than 70 mw,57- 173 the heat produced
will be conducted away through the detonator
without igniting the mercury fulminate. In
order to fire the detonator, energy must be sup-
plied at a rate faster than it can be safely con-
ducted away. The greater the energy dissipa-
tion is above this lower limit, the smaller the
time delay. Further, in order to waste as little
energy as possible in heat conduction, the total
energy should be supplied in as short a time as
possible. Thus, a constant energy dissipation of
1 w will fire the detonator in 1 msec. However,
because of the thermal inertia of the mercury
fulminate in contact with the Nichrome heater,
energy dissipation at the rate of 25 w is re-
quired to decrease the total time to 170 psec.

As a fuze may be used in the upper atmos-
phere, where the temperature is far below
freezing, the possible effect of low tempera-

] SECRET \

120

ELECTRONIC CONTROL SYSTEMS

tures on the action of the detonator is impor-
tant. The only measurable effect was that more
energy was required for detonation at lower
temperatures. This effect had been anticipated,
but even at —78 C only about 10 per cent more
energy was required.166

Leakage Resistance. Another characteristic
of interest is the resistance between the detona-
tor lead wires and its metal shell. This property
is important because of the possible firing of
the detonator by a voltage between the case and
one lead or by leakage otherwise affecting the
proper operation of the detonator circuit. The
minimum resistance measured in a group of
fifty detonators at ordinary temperature and
humidity was 50,000 megohms. When subjected
to a relative humidity of 95 per cent for 24
hours, the lowest resistance decreased to 12,000
megohms.24 The effect of leakage resistances of
these magnitudes upon the proper operation of
the detonator circuit can be neglected.

Specifications. A summary of the pertinent
operational characteristics of the detonator is
given in the specification212 which was used by
the Army for large-scale procurement. An ex-
tract follows.

The detonator shall function in an elapsed time not
exceeding 0.005 second with an electrical current of not
more than 0.175 ampere at 20°C, or with an electrical
current of not more than 0.225 ampere at — 15°C.
Eighty per cent or more of the detonators shall also
function in an elapsed time not exceeding 0.001 second,
and none over 0.003 second with the discharge from a
condenser of not more than 0.7 microfarad capacitance,
charged from a battery of not more than 75 volts
potential.

Some lots of detonators have had difficulty
in passing the 0.175-amp specification, which
is more severe than the other two ; conse-
quently, a recommendation that this current be
raised to 0.200 amp was made.

The Detonator Capacitor

The capacitor is used in the detonator circuit
as a very low impedance power source, which
must store enough energy to fire the detonator.
As far as the capacitor is concerned, the opera-
tion of the thyratron shorts a low resistance,
on the order of 18 ohms, across it. With the

minimum supply voltage specified as 125 v, the
nominal capacitance required to insure that the
detonator fires is 1 pf. A major requirement for
the capacitor is that it must be small enough
so that it does not occupy a disproportionate
amount of the very limited space in the fuze.
Although electrolytic condensers easily meet
the space requirements, they were rejected at
an early stage in the fuze development because
of their many faults.217 They deteriorate during
storage and then require a long forming period,
during which they pass excessive leakage cur-
rent and store very little energy. They cannot
withstand either low temperatures at high alti-
tudes, or high temperatures of the tropics. In
addition, all the energy stored in the electro-
static field is not immediately released, when
the capacitor is shorted, because of dielectric
hysteresis in the condenser. Paper condensers
are therefore used as being the most efficient
space utilizers which do not have these faults.
In order to eliminate the deleterious effect of
high-humidity conditions, the firing condenser
is metal-clad.

In all the generator-powered fuzes, except
those which use RC plate arming, the detonator
firing condenser also serves as the part of the
filter circuit of the power supply. The charac-
teristics which determine its effectiveness in
firing detonators are its capacitance, induct-
ance, and internal series resistance. The dielec-
tric absorption of a paper condenser is negli-
gible for a single discharge. The leakage re-
sistance is of only minor importance in a unit
which does not use RC arming, as long as it is
not so low that it presents an appreciable
powTer drain on the generator. A 5-megohm
leakage resistance will cause negligible drain on
the power supply.

The series resistance and the inductance of
the detonator firing capacitor are not measured
separately ; instead, a surge current test is
made, which is intended to determine how well
the capacitor will discharge through the deto-
nator and thyratron. The capacitor is charged
up to 135 v and then discharged through a
15-ohm resistor. The peak current is required
to be 7 amp. As either inductance or series re-
sistance in the condenser would lower the peak
current, this test gives an indication of the

SECRET

THE DETONATOR CIRCUIT

121

combined effect of both. Furthermore, as the
condenser is actually used this way, the test is
quite valid.

When measuring the peak surge current by
discharging the condenser through the 15-ohm
resistor, the impedance of the switch is in
series with the resistor. The spark that occurs
as the switch is closed represents a variable
impedance which limits the peak surge current.
It is, therefore, necessary to use a fast-acting
mercury switch, rather than an ordinary
switch.

The 7-amp limit was chosen because the bet-
ter capacitors were able to pass this test, and
the requirement allows a 100 per cent factor of
safety in firing the detonator. In Figure 41 is
shown the diminution in the peak surge cur-

Figure 41. Effect of series inductance on peak
surge current of detonator firing capacitor.

rent, due to the series inductance, when the
1.5 pf condenser used in the battery-powered
fuzes is tested. As the time lag is also of in-
terest, Figure 42 gives the time to peak current
as a function of the inductance.

The possible effect of the inductance in de-
creasing the energy dissipated in the resistive
component of the circuit was investigated. In
the actual circuit, the thyratron stops conduct-
ing when its potential difference falls below
about 20 v. The energy dissipated in the 15-ohm
resistor was, therefore, determined as a func-
tion of the inductance for a 1.5 pf capacitor dis-
charging from 135 to 20 v. Figure 43 shows that
even a 100-ph inductance would decrease the
energy dissipation less than 2 per cent. Thus, a
capacitor would have to be quite poor for its as-

sociated inductance to have an appreciable ef-
fect. As shown in Figure 41, with so large a
value of inductance the condenser would cause a
peak current of less than 7 amp, and would,
therefore, fail the peak surge current test. An

Figure 42. Effect of series inductance on time
to peak surge current of the detonator firing
capacitor.

inductance of 100 ph would decrease the peak
surge current to 6.4 amp. The effect of the in-
ductance on changing the time for the energy
discharge is negligible. As shown in Figure 44,
the time varies from 43 to 38 psec as the induct-
ance increases from 0 up to 100 ph.

The characteristics of capacitors may be
affected by the ambient temperature and hu-
midity, so that these factors must be taken into

INDUCTANCE (HENRIES)

Figure 43. Effect of series inductance (asso-
ciated with detonator firing capacitor) on energy
dissipated in resistance load.

account.166 The principal effect of a high rela-
tive humidity is to decrease the leakage resist-
ance of the capacitor. It was found necessary
to use metal-cased condensers in order to elimi-

122

ELECTRONIC CONTROL SYSTEMS

nate this difficulty. Temperature variations
affect both the capacitance and the leakage re-
sistance of the condenser. At low temperatures
the leakage resistance increases; at high tem-
peratures it decreases, showing that the dielec-
tric and the impregnating material have high
negative temperature coefficients of resistivity,
as might be expected. The capacitance de-
creases at low temperatures, so the condensers
used must pass the specifications at the lowest
operating temperature for the fuze. Moreover,
it is well known that both the capacitance and
leakage resistance change as a result of a tem-
perature cycle.217 Therefore, in order to deter-
mine how the condensers will react to different

Figure 44. Effect of series inductance on time
of capacitor discharge.

weather conditions, they must be temperature-
cycled several times. Data are taken during
each cycle, and the capacity and leakage resist-
ance are required to meet the specifications
throughout. Fairly large samples must be used
in these tests in order that the data represent
the condenser type, rather than just the sam-
ples tested. Another characteristic that is de-
termined during these temperature and hu-
midity tests relates to the mechanical strength
of the condenser assembly, particularly to as-
sure that the leads do not come out of the con-
denser. The temperature and humidity condi-
tions do not affect the operations of the con-
denser in the peak surge current test. This is

as expected, since the peak surge current is in-
dependent both of the capacity and of the leak-
age resistance if it is greater than 1,000 ohms.

The Thyratron

The thyratron is used in the variable-time
[VT] fuze as an extremely sensitive electronic
switch. Rather stringent requirements are
placed on the thyratron by the physical size of
the fuze, the available power supply, the char-
acteristics of the oscillator and amplifier sec-
tions, the detonator, and discharge condenser,
and the use to which the fuze is put.

Because of the limited volume of the fuze, it
was necessary to develop a thyratron8 occupy-
ing not more than % cu in. of space. Into this
space were fitted the components required to
give the thyratron the electric characteristics
required by the factors noted above. These re-
quirements were established as follows :

Low Power Consumption. In both battery
and generator powered fuzes the available
power is distinctly limited. (See Section 3.4.)
It was, therefore, necessary to design the fila-
ment for the lowest possible power consump-
tion consistent with the required life and surge
characteristics of the tube.

Critical Grid Voltage. An allowable range of
—2.1 ± 0.4 v for critical grid voltages was dic-
tated by the amount of bias voltage which
could be incorporated in the battery and the
available signal level from the oscillator-ampli-
fier section of the fuze.

Effective Critical Grid Voltage. The value
for critical grid voltage defined above is for d-c
operation. When the thyratron filament is pow-
ered by alternating current the critical grid
voltage is increased because the filament poten-
tial is negative during half of each cycle, and
the critical grid voltage is referred to the most
negative portion of the filament. As installed in
a fuze circuit the thyratron grid receives tran-
sient and ripple signals from the amplifier. The
phase of the ripple signal is usually such as to
reduce the effect of a-c ripple on the thyratron
filament. The highest negative bias at which
the thyratron will fire, under operating condi-
tions, is called the effective critical grid voltage.

SECRET

THE DETONATOR CIRCUIT

123

Stability. Supply voltages of generator-pow-
ered fuzes are generally higher than those pow-
ered by batteries. In addition, the battery volt-
ages change considerably with age and climatic
conditions. These factors necessitated a thyra-
tron whose critical grid voltage was as insensi-
tive as possible to changes in operating voltage
both from the standpoint of magnitude and the
ability of the grid to maintain control.

Surge Characteristics. The properties of the
detonator and discharge condenser, together
with the nature of the fuze application, deter-
mined the required surge characteristics. The
thyratron must be able to pass peak surge cur-
rents of the order of 7 amp in 0.001 sec after
triggering, in order to transmit the energy
from the discharge condenser to the detonator
in a time short enough to set up a high-order
detonation at the same point in space at which
the triggering signal was received, the speed of
missile being approximately 1,000 to 1,500 fps.

Leakage and Grid Current. Both leakage be-
tween plate and grid and grid current contribute
to unstable critical grid voltages and therefore
must be minimized. Where RC arming is used
in addition to mechanical arming, leakage be-
tween plate and filament is important and must
also be minimized.

Microphonics. Because of the multitudinous
vibrations and shocks to which the tube is sub-
jected in operation it was necessary to have the
thyratron mechanically strong, so that it would
not operate prematurely.

Life. Although the fuze itself needs to oper-
ate only once, a certain amount of testing is
required prior to use. Since it appeared that
one way to obtain all the required electric char-
acteristics in so small a tube was to sacrifice
greatly in the time of useful operation, it was
necessary to preserve sufficient reserve to guar-
antee proper operation after the testing. Fur-
ther limitations are discussed in Section 3.1.4.

The first step in obtaining a suitable thyra-
tron was to examine the existing types of small
tubes. Tests were made on the Bell Telephone
Laboratories type 1278 GY-2 (see Figure 45),
the General Electric miniature thyratron and
the Sylvania type SN-738.6 The 1278 GY-2 was
soon eliminated as a possibility for several rea-
sons. Although the critical grid voltage of this

type was too high and spread over too great a
range, the principal objection was the manner
of construction of the tube itself. The geometry
of the tube was such as to make it susceptible
to external fields, making it impossible to use
the tube in closely packed assemblies without
careful external shielding. This was undesir-
able because of the premium on space.

The GE miniature thyratron8 had the dis-
advantages of excessive size, susceptibility to
external leakage caused by handling, and a
limitation of the number of times it could be
surged. Because of its manner of construction,
it had a distinct advantage in that the critical
grid voltage could be very closely controlled,
both as to magnitude and stability. It was,
therefore, decided that the size should be re-
duced and an attempt made to reduce the sus-
ceptibility to leakage and improve the surge
characteristics.

The Sylvania SN-738 had been developed for
Section T and, in general, was found to have
the proper characteristics with the exception
that it was designed to operate at low voltage
and exhibited poor performance at the higher
voltages used in Division 4 fuzes. It was de-
cided that this tube should be redesigned to
operate at the higher voltages and also to elimi-
nate certain classified features peculiar to the
original purpose for which it was designed.

The redesigned GE and Sylvania thyratrons
were known as the microthyratron and SA-782
respectively (see Figure 45). The microthyra-
tron preserved all the advantages of close grid
control found in the miniature thyratron and
largely eliminated the problem of leakage by
virtue of a special lacquer coating over the sur-
face of the tube. The microthyratron is essen-
tially a cold cathode tube, having a filament
supplying only enough emission to initiate the
discharge which immediately transfers to an
anodized aluminum spotting tab. This tube was
finally abandoned, because it was not possible
to maintain the filament emission at the proper
value over the range of filament voltages to
which the tube was subjected, nor was it pos-
sible to obtain a spotting tab which would stand
the punishment of repeated surging of the
thyratron.

The Sylvania thyratron in its final form as

124

ELECTRONIC CONTROL SYSTEMS

the SA-782B (shown in breakdown in Figure
46), meets all of the requirements outlined
above. The outstanding features in the design
of this tube were the introduction of an addi-

Figure 45. Various thyratrons developed and
considered for use in radio proximity fuzes.
From left to right they are : GE microthyratron,
BTL 1278 GY-2, GE version of Sylvania

thyratron, and Sylvania SA-782B. The latter
tube was also known by the NDRC designation
NS-4 and the Signal Corps designation 2D29.

tional grid between the control grid and the
anode and an auxiliary shield around the fila-
ment above the top mica. The additional grid
is connected to the negative leg of the filament
and makes possible a lower critical grid voltage
than is otherwise consistent with the geometry
of the tube and also controls the spread of criti-
cal grid voltage from tube to tube. The auxil-
iary shield is connected to the control grid and
makes stable operation possible at higher than
normal voltages. This shield was necessitated
by the fact that the emitting portion of the fila-
ment sometimes extends above the top mica
and at sufficiently high anode voltages causes
an arc-over to the anode because of the ab-
sence of grid control in that part of the
tube.8, 67, 189, 202, 213

Circuit Operation

The sequence of operations in the fuze after
the projectile is released is such that it can
function as soon as arming is completed. Upon
release the generator propeller starts to turn,
and it reaches the equilibrium rotational veloc-
ity, due to its speed through the air, in 5 to 6

revolutions. The tube filaments warm up with-
in 0.4 sec,11 and both oscillator and amplifier
are in operation. The B voltage has already
reached its steady-state value. However, the
fuze cannot function, because the electric deto-
nator is not yet connected to the firing circuit.
The details of mechanical arming are covered
in Chapter 4 ; here it will suffice to state that a
preset number of turns of the fuze propeller is
required before the detonator makes electric
contact. Until this time, the detonator is sepa-
rated from the tetryl booster by a %- in. thick
brass plate. Upon arming, the detonator bridge
wire is connected between the detonator firing
condenser and the thyratron plate. Except
when using RC arming (see Section 3.3.6), the
firing condenser is also the output filter con-
denser and is already charged up to the operat-
ing potential of about 140 v. At electric arm-
ing, therefore, a 140-v positive pulse is applied
to the thyratron plate. A part of this arming
pulse appears on the thyratron grid in the ratio

f n

Figure 46. Breakdown of thyratron showing
electrode SA-782B structure.

of the grid-to-filament impedance to plate-to-
grid impedance. In order to prevent this grid
pulse from firing the thyratron, the grid-to-fila-
ment impedance is reduced by connecting a con-
denser from grid to ground (shown in Figure
39). In the early fuzes a 50-qpf condenser was
used and was quite satisfactory. In later pro-
duction fuzes, when a 500-ppf condenser was
used in order to get proper amplifier shaping,

THE DETONATOR CIRCUIT

125

an additional safety factor was provided. The
presence of this arming pulse places an addi-
tional requirement on the thyratron; the leak-
age resistance from the plate to the grid must
be quite large, on the order of 1,000 megohms.
At any time after arming, a sufficiently large
signal, 3 to 4 v positive, impressed on the thyra-
tron grid by the amplifier, will fire the thyra-
tron ; and the condenser will discharge through
the tube and the detonator, initiating the ex-
plosion.

Just as there is a time lag for a signal to
travel through the amplifier,54 so there is a time
lag between the incidence of the firing signal
at the thyratron grid and the explosion of the
detonator.25 This time lag is on the order of
1 msec and is almost entirely due to the de-
tonator. The delay due to the thyratron is usu-
ally less than 150 psec189 and that due to the
condenser is less than 60 psec. One millijoule
of energy dissipated in the detonator in 1 msec
will set it off ; in order to decrease this time
lag, it is necessary to dissipate more energy,
faster, in the detonator. For example, this time
delay can be reduced to 200 psec by dissipating
3.6 milli joules in it in this time.58

Component Values. To determine the effi-
ciency of a given capacitance in firing a de-
tonator, a condenser of the given size is used to
fire detonators through thyratrons in the deto-
nator circuit. The condenser potential is ini-
tially too low to fire the detonator when the
thyratron fires, and it is gradually increased
until the detonator does function.25 By using
several thyratrons, detonators, and condensers,
the spread in firing voltage due to variations in
these components is determined. These data are
important to determine the minimum capaci-
tance that may be used to fire the detonator.

Tests were made on the detonator circuit,
with all components, beyond the extreme tem-
perature limits foreseen for actual operation,
namely, —78 C and 60 C, to determine the
effects of extreme temperatures. The power
supply specifications permit a minimum B volt-
age of 125 v under normal conditions ; however,
at low temperature the efficiency of the sele-
nium-button rectifier assembly decreases. At
—40 C, the normal output of 135 v will be only
106 v 1 sec after launching,69 at which time a

rocket fuze should be ready to function. This B
voltage was used in some of the tests, as was
the minimum expected A voltage, 1.17 v rms.178
The filament series resistor lowered the thyra-
tron filament voltage still further. It should be
noted that this lowered filament voltage actu-
ally has a beneficial effect. Although the de-
creased filament emission increases the thyra-
tron time delay, this time is still sufficiently
small. In addition, because of the lower fila-
ment operating temperature, the filament re-
sistance is lower. This decreased series resist-
ance permits a larger fraction of the condenser
energy to be dissipated in the detonator. After
making allowances for all the other factors
which influence the operation of the detonator
circuit, it was found that the minimum capaci-
tance which would fire any good detonator,
using any thyratron which passed the tube
tests, was 0.96 pf.58 It became apparent in these
tests that the largest single variable in deter-
mining the firing capacitance and voltage needed
was the thyratron, although all the tubes used
had passed the tube tests.

The thyratron is used as a low-impedance
switch to permit the energy stored in the con-
denser to be dissipated in the detonator. About
40 per cent of the condenser energy is actually
transferred to the detonator.6

The minimum capacitance value can be fur-
ther lowered by a selection test for thyratrons58
(this technique was not used in production).
Such a test measures the efficiency of the thyra-
tron in permitting energy transfer. Rejection
of less than 5 per cent of the thyratrons which
pass the thyratron specification tests would
lower the minimum value of the thyratron-
firing capacitor to 0.87 pf. With further divi-
sion of the tubes into two approximately equal
groups, the better group would permit the use
of an 0.5 pf firing capacitor in fuzes where the
space requirements are critical. The other
group of thyratrons could be used in the larger
fuzes, where the space requirements are not so
stringent.58

3,3 6 Electric (RC) Arming

The use of a resistance-capacitor network to
delay arming provides a delay in electric arm-

SECRET

126

ELECTRONIC CONTROL SYSTEMS

ing (RC) after mechanical arming has oc-
curred.58 It consists of placing a large resistor,
on the order of a megohm, between the single
filter capacitor used with RC arming and the
detonator firing capacitor. At mechanical arm-
ing the detonator firing capacitor starts
charging. The fuze cannot function until the
capacitor potential is high enough to set off the
detonator if the thyratron should fire. The de-
tonator firing capacitor and the arming resistor
are shorted by a large resistor, so that there is
no charge on the capacitor before mechanical
arming occurs.181 A diagram of the circuit as
used in the T-171 is shown in Figure 47. The
time delay before electric arming is propor-
tional to the size of the arming resistor. Time
delays up to approximately 8 sec can be

Figure 47. RC arming circuit of T-171.

achieved by proper choice of the resistance. An
upper limit is set by the leakage resistance of
the capacitor,163 in comparison with the series
resistor used. Also of importance is the accom-
panying elimination of the arming pulse when
RC arming is used.

The data on the potential necessary on a par-
ticular capacitance to fire the detonator are of
primary importance when RC arming is used,
since they permit the calculation of the time
interval between mechanical arming and elec-
tric arming.56 This period is the time taken for
the capacitance to charge to the firing voltage,
and it is a function of the supply voltage, the
capacitance, the arming resistance, and of the

particular thyratron and detonator. In making
arming calculations, the median voltage re-
quired to fire a detonator through a thyratron
using a particular capacitance is used, rather
than the average voltage. The median voltage
is the voltage which will fire half the detona-
tors, and it represents the firing data much bet-
ter than does the average voltage, which is
unduly influenced by extreme values due to
atypical components. The data on firing detona-
tors using different capacitances are summar-
ized in the following table.

Median voltage required to fire detonator.
Capacitance Median voltage

0.3

0.37

0.5

0.75

1.0

1.335

1.555

1.7

96.5

93.5

84.0

74.0
69.2

63.0

60.0
59.0

The above values are plotted in Figure 48 as
the simplest method of averaging all the data,
as well as permitting determination of the
median voltages required for capacitances not
used in the tests.

Another method for averaging the data is to
fit a least squares curve to the points on the
graph, with the added advantage of permitting
algebraic computations with the resultant equa-
tion. As the data are the voltages required for
various capacitances to fire the median det-
onator, just enough energy is dissipated in the
detonator to fire it. The least squares curve,
therefore, indicates a constant energy dissipa-
tion in the resistive portion of the detonator
circuit. It also shows the condenser potential at
detonation to be very nearly equal to the con-
stant voltage drop Vf of both the gas in the
thyratron when it is conducting and the contact
potentials in the tube.

The condenser, initially charged to a poten-
tial V, stores an amount of energy

Wi = \CV\ (33)

After detonation, the residual energy on the
condenser is

Wf = iCV /. (34)

After the thyratron starts conducting, before
the arc starts, the condenser potential drops to

THE DETONATOR CIRCUIT

127

V a, and the energy loss, almost all of which is
dissipated across the gas in the tube, is

Wa = \CV2 - iCVa 2. (35)

From the initiation of the arc to its extinction,
the constant potential Vp due to both the tube
contact potentials and the gas, causes an energy
loss of

Wt = QVf = ( CVa - CVf) Vf. (36)

The energy dissipated in the resistive portion
of the circuit is

W = Wi - Wf - Wa - Wtt

W = iCV 2 - \CVS2 - iC(V2 - Va2)

- (77,(7* - Vf). (37)

Solving,

W = \C {V a - Vf)2. (38)

As the experimental datum is V,

W = \C{V - V + 7* - Vf)2,
or

W = J C[V - (V - 7* + Vf)]2. (39)

The least squares fit of this equation gives val-
ues of W = 0.67 millijoule, and V — Va + Vf =
31.6 v. Vf is about 18 v, so that V — Va , = 13.6 v,
which is approximately the condenser potential
drop before the thyratron arc strikes. This
value has been verified by oscilloscope measure-
ments.58 As both these voltage drops are due to
the thyratron, the equation may be rewritten,
W = YzCiV — Vt) 2. The current during the ini-
tial 13.6-v drop is comparatively small, and
most of this energy is dissipated in the tube in
starting the arc with only a negligible amount
of it dissipated in the detonator. It should be
noted that any energy stored in the magnetic
field of the circuit inductance at peak current
has been dissipated in the resistive portion of
the circuit by the time conduction stops.

Several points from equation (39) are
plotted as the circles in Figure 48 using the
above value for the bracket term, but the com-
plete graph was not drawn because it is almost
indistinguishable from the curve drawn
through the experimental points. The excellent
fit to the eight experimental points, using a
two-parameter equation, validates the form of
the equation as being the type to which the data

conform. As this equation is approximately a
straight line on log-log graph paper, the ex-
perimental points were plotted on such paper
as in Figure 49. The curve which fits them best
is the straight line drawn through them. The
points from the theoretical curve are also
plotted in Figure 49.

Using the experimental data for the median
voltages to fire detonators through thyratrons
with various capacitances and assuming an av-
erage unit B supply of 140 v, the equation

Figure 48. Median values of voltage on ca-
pacitors necessary to fire detonators through
thyratrons for various values of capacitance.

t/R — —C In (1 — V/V0) was solved for a
value of t/R to correspond to each pair of
capacitance and firing voltage values.58 The t in
this equation is the time delay in charging a
capacitance C to a voltage V through a resist-
ance R, using a supply voltage 70. These points,
which represent data for 250 thyratrons, are
plotted in Figure 50, where the time in seconds
divided by the resistance in megohms is plotted
against the capacitance in microfarads. This
curve can be used to determine the median elec-
tric arming time in fuzes using RC arming for
any resistance value, and the curve extends be-

-SECRET

128

ELECTRONIC CONTROL SYSTEMS

yond the values of capacitance that have been
used in the fuzes to date.

The spread in arming times around the
median values due to variations in the compo-
nents has been represented in Figure 51, which
can be used to find the actual time in which a
given percentage of the fuzes with a certain
median arming time will be electrically
armed.36 In making the calculations, it was
assumed that both the resistors and the con-
densers were within 10 per cent of their nomi-
nal values, with every value in this range
equally probable. The supply voltage was as-
sumed to vary between 125 and 160 v in a para-
bolic probability distribution, with the center
value three times as likely as the extreme val-
ues. The distribution of detonator firing volt-
ages, using a given capacitance, was experi-
mentally determined, using many thyratrons

CAPACITANCE (MICROFARADS)

Figure 49. Median values of voltage on ca-
pacitors necessary to fire detonators through
thyratrons for various values of capacitance on
log-log scale.

and detonators. It was also assumed that the
leakage resistance of the condenser was high
enough so that it would not affect the condenser
potential appreciably. This assumption requires
that the leakage resistance be greater than 40
megohms when a 2-megohm arming resistor is
used, and greater than 25 megohms when a
1-megohm arming resistor is used.163 As previ-

ously noted, the thyratron is responsible for
most of the spread in arming time.

Dumping. The use of RC arming is of de-
cided advantage when the tactical use of the

Figure 50. Median arming time in RC circuits
as related to resistance and capacitance values.

fuze is such that spurious signals of firing mag-
nitude are possible shortly after mechanical
arming.84 For example, as used on rockets, the
phenomenon known as afterburning, whereby,
after the main burning of the propellant is
over, additional slivers of propellant ignite ;
the rocket expels quantities of luminous gas
and produces several random, intermittent sig-
nals of several times firing magnitude. With
RC arming,58 if a firing signal is impressed on
the thyratron grid before the plate potential
has reached about 38 v, a low-current discharge
will start through the thyratron, gradually dis-
charging the detonator firing condenser. When
the signal is over, the thyratron grid regains
control of the tube, the discharge stops, and
the condenser starts to charge up again. If the
firing signal occurs after the thyratron plate
potential has exceeded 38 v but before the con-
denser has stored enough energy to fire the
particular detonator through the particular
thyratron, an arc discharge takes place. The
condenser potential drops to about 18 v, after
which the grid regains control and the con-
denser starts to charge up again. This occur-

SECRET

THE DETONATOR CIRCUIT

129

rence is known as “dumping.” In the arc dis-
charge, the thyratron begins to conduct within
10 |isec after the signal is impressed on the
grid. The effective thyratron impedance de-
creases to about 10 ohms, thus permitting a
very high peak surge current in the neighbor-
hood of 6 amp. The discharge is over within
about 50 psec. When used with rockets which
suffer badly from afterburning, this phenome-
non of dumping permits the use of radio prox-
imity fuzes without undue incidence of early
functions. At the same time, it permits the fuze

Figure 51. Distribution of RC arming times in
terms of the median arming time.

to function properly within the shortest per-
missible time after burning has stopped. If
afterburning is severe, the thyratron may dump
several times before electric arming occurs. In
this way, the use of RC arming provides insur-
ance against premature fuze functions which
might otherwise occur shortly after mechanical
arming.

Arming Pulse Protection. In some of the
fuzes, such as those designed for use on mortar
shells, mechanical arming causes a pulse to
originate in the oscillator. This voltage pulse
cannot be protected against by means of the

condenser in the thyratron grid circuit, which
only decreases the effect of the thyratron plate
pulse. The use of RC arming does eliminate the
effect of such arming pulses, but it requires the
use of a detonator firing condenser in addition
to the filter condenser. This necessitates the
allocation of space to two large components
in the fuzes where the space requirements are
most critical. As there are no afterburning
problems with a mortar shell, an electric arm-
ing system which would eliminate the effect of
the arming pulse is all that is necessary. Such
a system has been developed which also permits
the use of the same condenser for filtering the
rectifier output and for firing the condenser.
The scheme consists in having large negative
voltage on both the thyratron and amplifier
grids, in addition to the normal grid bias, be-
fore mechanical arming. This C voltage is large
enough to bias the amplifier tube beyond cutoff.
At mechanical arming the additional C bias is
eliminated, permitting the amplifier to start
functioning. The time constants in the fuze cir-
cuits are so adjusted that the arming pulse
from the oscillator is harmlessly over before
the amplifier has reached its normal operating
point and before the thyratron grid bias has
reached its normal value.

Other Arming Methods. Several other elec-
tric arming methods have been used during the
development period. One method of eliminating
the pulse which occurs at arming is to use a
system which does not change the potential at
any point in the fuze. Two of the systems in-
volve only the detonator, which must fire in
order to initiate the explosive in the projectile.
The first one consisted of having a wire short-
ing the detonator before arming, so that all
the circuits would have reached equilibrium by
the time the detonator was unshorted at arm-
ing. This method was abandoned because it did
not assure perfect safety, because of the possi-
bility of the shorting contact opening prema-
turely due to vibration or shock. Moreover,
there was difficulty in keeping the thyratron in
the nonconducting state at arming if a firing
signal reached the thyratron grid before arm-
ing. This method was thus used primarily as a
safety device rather than as an arming method.
It was replaced by a system which used a 10-

130

ELECTRONIC CONTROL SYSTEMS

megohm resistor in series with the detonator.
The resistor was shorted at arming. If the
thyratron fired before arming, the 10-megohm
resistor limited the current flowing through
the detonator to a very small value, on the
order of 12 qa, so that the heat generated in
the detonator could be safely dissipated. This
current is so small that, even if all the heat
energy produced on the detonator bridge wire
remained there, it would take several hours to
initiate the explosion. The factor of safety in-
volved is enormous. The minimum constant
current which will fire the detonator is about
90 ma,173 and the detonator will function in
about 5 msec. The heat produced by any small
current can be safely conducted away57 through
the mercury fulminate, which initiates the ex-
plosion in the detonator. This arrangement is
much safer because it is possible to place the
contacts so that accidental operation due to
vibration or shock is impossible. It also per-
mitted the inclusion of a self-quenching feature,
if the thyratron fired before arming, by con-
necting a condenser between the thyratron plate
and ground. This system was the precursor of
RC arming.

Other arming systems tried also permit the
thyratron plate to be at its operating potential
and rely on preventing the thyratron from
functioning and thus preventing detonation. In
one of these systems a large negative bias is
impressed on the thyratron grid, which is re-
moved at arming. In order to improve this
method, an RC time delay was added at the
thyratron grid to prevent functioning immedi-
ately after arming had occurred.

Another arming system was used in some of
the experimental fuzes developed at the Bell
Telephone Laboratories.204 The pentode screen
resistor is connected to B-f- through the thyra-
tron plate network. Until mechanical arming,
when the thyratron plate circuit is closed, the
pentode screen potential is slightly negative,
because it assumes an equilibrium potential due
to the electron current in the tube. The gain
of the amplifier with this screen potential is
virtually zero and consequently ho signals are
passed to the thyratron. After mechanical arm-
ing there is a time delay during which the
screen by-pass condenser is charged, before the

amplifier gain reaches its normal value. The
voltage transients due to the arming process
are over before the amplifier is operative, so
that none of the arming transients can cause
the fuze to function prematurely. Furthermore,
as the pentode becomes operative, its plate po-
tential drops from B+ to its normal operating
potential and transmits a large negative pulse
of short duration to the thyratron grid. Still
another method, the principle of which is still
in use, is to operate the thyratron filament at
a lower voltage than the other tube filaments,
thus increasing the time delay before the thyra-
tron can function beyond that of the rest of the
fuze. In this way, most pulses are over before
the thyratron becomes operable. The lower
thyratron filament voltage is obtained by in-
creasing the resistor in series with the thyra-
tron filament. As used on generator-powered
fuzes, the net effect is that a higher rotational
speed is required for the thyratron to be able
to function than for the other tubes.

3‘3’7 Safety Features

This part of the report deals only with the
electric safety features, many of which have
already been discussed in the preceding parts
of this section dealing with electric arming.
Actually, the two are very closely related. The
primary method for assuring that the fuzes
will be entirely safe in the unarmed position is
to make certain that no current can flow through
the detonator heater wire. This may be done
either by having the detonator open-circuited
before arming, which is the method used in all
production fuzes, or by having the detonator
short-circuited before arming. By relaxing the
no-current requirement to permit a minute cur-
rent, the use of the 10-megohm resistor in series
with the detonator might also be included in
this classification.

The other possible electric safety feature is to
prevent the thyratron from firing prematurely
and thus prevent detonation. The methods used
are RC arming in either the plate circuit or
the grid circuit of the thyratron. An additional
safety feature common to all generator-powered
fuzes is that, as long as the generator is not

SECRET

POWER SUPPLIES

131

turning, there is no electric energy available in
the fuze. Hence, the detonator cannot fire. Of
course, this last feature depends on the exist-
ence of a leakage path across the detonator
firing condenser, so that the condenser will not
still be charged from the previous occasion
when the generator was functioning.

3 38 Self-Destruction

Electric self-destruction [SD] was used in
many of the battery-powered fuzes (T-5) . When
these fuzes are used as antiaircraft weapons
over friendly territory, it is necessary to pre-
vent them from exploding on ground approach

last factor caused the largest variation in the
time delay. If the neon gas is relatively un-
ionized the striking potential may be increased
as much as 30 per cent above its normal value.
Methods used for keeping the gas sufficiently
ionized to minimize variations in striking volt-
age were to channel light to the neon tube
through a Lucite window, and thus ionize the
gas photoelectrically, and to place a little radio-
active material on the tube envelope. It was
found that cosmic radiation did not keep the
gas sufficiently ionized.

POWER SUPPLIES8

DETONATOR

Figure 52. Self-destruction circuit used in T-5
fuzes.

and inflicting casualties. The method used was
to explode the fuze from 6 to 11 sec after the
missile was launched, if it had not already
functioned. This time limit was long enough
so that, if the projectile were going to function
on a target in combat, it would already have
done so. The SD device consisted of an RC time-
delay circuit (shown in Figure 52) at the thy-
ratron grid, which was connected to a neon
tube. The neon tube used was the General Elec-
tric Company [GE] NE-23, a 1/25-w lamp
in a T-2 bottle. When the condenser potential
reached the striking potential of the neon tube,
the condenser discharged through the neon
tube, and thus fired the thyratron, which in
turn set off the detonator. The time delay in
this circuit is a function of the resistance, the
capacitance, and the battery voltage, as well as
of the neon tube striking potential. In fact, this

Requirements

The radio-type proximity fuze requires an
electric power supply which provides filament,
plate, and grid bias voltages to the electronic
system and which ultimately delivers a current
surge to the electric detonator upon actuation
of the fuze. The power supply has, therefore,
a position of prime functional importance. The
quality of overall fuze performance cannot ex-
ceed that of the power supply. Much effort has
been directed toward the design of a power
supply meeting the varied requirements pecul-
iar to proximity fuze operation.

Because of the urgency of the program it
was not practical to follow an optimum pro-
cedure of setting up rigid functional specifica-
tions for each part of the device and aiming all
development to these prescribed requirements.
Instead all parts of the device were under
simultaneous development toward rather elastic
specifications. Progress of one phase invariably
produced a concomitant change in the mini-
mum requirements of another phase.

The initial goal was to realize rapidly a fuze
design which would provide consistent function
even if expediency required compromise of the
ultimate qualities which were visualized for
the fuzes. Subsequent redesign was relied on
to introduce improved versatility, performance

£ This section was written by J. G. Reid, Jr., of the
Ordnance Development Division of the National Bureau
of Standards.

SECRET

132

ELECTRONIC CONTROL SYSTEMS

quality, ruggedness, and simplicity of construc-
tion. This general philosophy held not only in
the design of the unit as a whole but also in the
design of subassemblies, such as the power sup-
ply. The following general specifications for the
power supply developed during progress of the
work. They represent partly initial ideas and
partly modifications imposed by service con-
siderations.

The primary requirement upon the power
supply system is to furnish adequate operating
power to the electronic system. No compromise
can be made on this point. Accordingly, mini-
mum requirements for A, B, and C supply were
established on the following bases:

Filament Supply. Circuit designs utilized
electron tubes of maximum 1.5-v nominal fila-
ment operation. Other tube types requiring
lower filament voltages were adapted to 1.5-v
supply by use of the proper filament dropping
resistors. At normal filament voltage various
tube types drew currents ranging from 70 to
220 ma. The tube complements of various fuzes
had filament requirements ranging from 450 to
750 ma, equivalent to 0.6 to 1.1 w. The fore-
going are d-c values or rms a-c values.

Plate Supply. Plate current demands of the
electronic systems of various fuzes were rela-
tively uniform. Radio-frequency oscillators
drew an average of 12 ma at 135 v with a
spread of about ±2 ma. Amplifiers in all cases
drew less than 0.5 ma at 135 v. Thus, total plate
circuit requirements lay between 2 and 3 w
supplied at 135 to 150 v.

Bias Supply. Negative grid-bias voltages of
about 6 v to the thyratron and 1.5 v to the
amplifier were also required of the power sup-
ply. These voltages were applied to such high-
resistance loads that the power involved was
negligible.

Detonator Firing. Type BS-4 and BS-5 det-
onators (see Section 3.3) were used in all fuzes
developed by Division 4. Actuation of either of
these required an internal dissipation of 1 mil-
lijoule of electric energy within the duration
of 1 msec, or great energy over a longer period.
A minimum of approximately 5 milli joules was
required from the power supply within 1 or
2 msec, the excess supplying energy losses else-
where in the firing circuit and insuring con-

sistent function of the detonator. As explained
in Section 3.3, it was found expedient to draw
this energy from a noninductive capacitor, 1 mf
or more in capacitance and charged to the volt-
age of the plate supply, i.e., at least 100 v. This
corresponded to a current peak in excess of
6 amp. The charging of the detonator-firing ca-
pacitor represented a negligible load on the
power supply. However in power sources where
the terminal voltage deteriorated with time the
lower voltage limit required for the capacitor
became very important.

Life. It was required that the power supply
maintain adequate voltage and power over an
operating period of at least 1 min. Testing re-
quirements presented additional demands and
when the fuze power supply was used for test
purposes (in production) approximately 10 min
additional life was essential.

Indefinite shelf life of the power supply was
desired but this requirement was waived in
some of the earlier battery-powered fuzes.

The power supply was required to operate
compatibly with the electronic system of the
fuze, not only in supplying operating voltages
sufficiently constant and noise free, but also in
not introducing excessive electric or mechanical
disturbance by its own operation.

Stability. Relatively small fluctuation in the
B supply voltage could cause fuze malfunction.
If the noise frequency coincided approximately
with that of amplifier peak gain, 0.03 per cent
magnitude was sufficient amplitude. Random
fluctuations of nearly twice this value were
permissible. Fluctuation in the A supply of
about the same absolute amplitude (or 100
times greater in percentage) was tolerable. In
summary, the voltages supplied had to be es-
sentially noise free. The precise value of toler-
able noise depended on the character of the
noise. This has been discussed in Sections 3.1
and 3.2.

It was further necessary that the power
supply, if it contained any moving parts, intro-
duce a minimum of mechanical disturbance to
the electronic system. The maximum tolerable
amplitude of vibration cannot be specified. It
depends necessarily on the type of construction
and the care that has been taken in designing
both circuits and components. It can be noted

POWER SUPPLIES

133

that the use of high-speed rotating or vibrating
parts was recognized as a possible source of
mechanical disturbance and consequent awk-
ward design problems.

Ambient Conditions. The range of ambient
conditions for fuze operation was made pro-
gressively broader. For the majority of the
fuzes developed, satisfactory operation was re-
quired in the temperature range from —40 to
+60 C, and from 0 to 100 per cent relative
humidity. This overall requirement was im-
posed also on the power supply. Further it was
desirable that proper fuze operation be obtained
after the unpackaged fuze had been maintained
under any such conditions for as long as
24 hours prior to use.

The fuze, including power supply, was, with
suitable moistureproof packaging, to have an
indefinite shelf life under the same ambient
conditions as for operation.

Ruggedness. Minimum requirements for the
mechanical strength and ruggedness of the
power supply were identical with those for
the fuze unit as a whole. In use it should per-
form satisfactorily after setback (for rocket
application, 250# maximum; for mortar appli-
cation, 10,000# maximum). No unusual precau-
tions should be required in fuzing or in the
handling of fuzed projectiles beyond the care
exercised in handling ordinary point-detonating
fuzes. The packaged bulk lots of fuzes should
be capable of withstanding the rough handling
such items customarily encountered in field
storage and delivery.

Size. With regard to physical design it
was desirable that the power supply be small
and of proper shape to match the remaining
sub-assemblies of the fuze. The particular
volume corresponding to “small” was progres-
sively reduced as the program advanced. In
first designs the power supply was allotted
about 12 cu in. in the shape of a cylinder 2.5 in.
outside diameter and 2.4 in. long. In later
models the total volume was reduced to less
than half this value.

It was a prime requisite that the power sup-
ply be adaptable to quantity production, at least
with regard to simplicity of construction. Econ-
omy of material was not a basic consideration
except in the case of strategic war materials.

3’4'2 Survey of Possible Sources of Power

In the development of a power supply to meet
the foregoing specifications, serious consider-
ation was reduced to two general types, those
using batteries to derive the required electric
energy from a chemical source and those em-
ploying electric generators driven by an input
of mechanical energy.

The electric demand upon the supply was a
maximum of 3 w for the plate circuit and 1 w
for the filaments. With a design excess of 25 per
cent this indicated 5 w for the power supply
output. The service life was to be a maximum
of 60 sec. Thus, for design purposes 300 j was
taken as the energy requirement upon a power
supply of battery or of generator type.

The space available for a battery-type power
supply was of the order of 10 cu in. or 160 cu
cm. This permitted a battery mass of about
250 g. The energy density of ordinary batteries
ranged from about 10 to 30 whr per kilogram,
i.e., 40 to 120 j per gram. Thus a minimum
energy content of 10,000 j could be realized
from a battery of acceptable dimensions. The
total service demand of 300 j then represented
only 3 per cent of the minimum expected total
energy within a 60-sec period. This was clearly
a tolerable class of battery service.

Dry Batteries. The ordinary carbon, zinc,
ammonium chloride dry battery presented itself
as the most readily available source of energy.
It was the consensus, however, that consider-
ations of delayed service and performance at
low temperatures would significantly limit its
usefulness.2 Although these dry batteries could
initially fulfill a pressing need with minimum
delay, it seemed doubtful that they could be
sufficiently improved as to be rendered com-
pletely satisfactory. The ultimate solution
would lie in some basically superior power
supply system.

Reserve Batteries. A reserve battery, in which
the electrolyte was introduced just prior to use,
could obviously meet the requirements of an
indefinitely delayed service period. Further-
more, a battery of this type exhibiting excellent
low-temperature performance had been devel-
oped by the Electrochemical Section of the
National Bureau of Standards.3 It used lead

SECRET

134

ELECTRONIC CONTROL SYSTEMS

oxide electrodes and a perchloric acid electro-
lyte. Some problems persisted, however, with
regard to the introduction of the electrolyte.

Two alternatives existed here : electrolyte
could be carried, separately packaged, within
the battery assembly to be actively introduced
upon application of some force due to projectile
acceleration; or electrolyte could be packaged
entirely separately from the battery and fuze,
for introduction shortly before the insertion of
the fuze into the projectile. The former required
the development of novel battery designs to
insure the rapid and proper distribution of
electrolyte and the maintenance of operating,
electrically discrete cells. The latter necessi-
tated the development of special filling equip-
ment and techniques which would be suitable
for the uncontrolled conditions of field use.
Also, the battery once made active could have
a shelf life of approximately one day. After
this period it would become useless, since re-
charging had proved unfeasible. Because of
these disadvantages most engineering effort on
electrolyte introduction was directed toward
setback actuated systems. Although this method
appeared possible in rocket and mortar appli-
cations, it was recognized that some externally
triggered force would have to be provided in
the case of bomb application where no setback
would be present upon release.

The principal difficulties expected in the de-
sign of a reserve battery supply system lay in
providing a high-voltage section consisting of
a multiplicity of series connected cells using
electrolyte from a common source. Here the
problems of rapid thorough distribution of
electrolyte and its subsequent retention with-
out intermittent short circuiting became acute.

Battery Vibrator. Consideration was given
the use of a vibrator high-voltage supply. Here
a single low-voltage high-capacity battery could
supply both filaments and the vibrator input.
The design study of such a system was under-
taken by the Washington Institute of Tech-
nology. Although their work indicated the gen-
eral feasibility of this system,203 it was not
carried to the point of achieving a battery,
vibrator, transformer, rectifier, and filter of
the requisite small size.

Generator. The optimum solution of the

power supply problem appeared to lie in the
use of a mechanically driven rotary generator.
Such a system offered several advantages, as
follows :

1. An indefinite delay prior to use would not
adversely affect its performance.

2. The generator would not be appreciably
affected by temperature extremes.

3. The entire fuze unit, including power sup-
ply, could be shipped into the field assembled
for use. No final assembly and test just prior
to use would be required.

4. The rotating system of the generator could
be coupled to suitable gearing to provide a me-
chanical arming and SD feature, if desired.

5. If the generator were wind-driven by the
flight of the projectile, a considerable additional
safety would accrue, since the power supply
would be inert prior to the period of service.

Since the generator served merely as a con-
verter rather than a storage source of energy,
it could be quite small. A volume of 2 to 3 cu in.
would suffice for an alternator of requisite
power. The additionally available space of 8 or
9 cu in. could accommodate the rectifier-filter
system and the prime mover for driving the
generator.

Prime Movers for Generators. Two basically
different conceptions of the prime mover were
apparent: (1) a storage system which received
an initial charge of mechanical energy prior
to the service period, and (2) a wind-driven
system continuously drawing energy from the
windstream during the flight of the projectile.
A rotating flywheel, a stressed spring, or a vol-
ume of compressed gas represent mechanical
systems of the storage type. In any case a
mechanical input of approximately twice the
electrical requirements would be necessary,
since efficiency of little more than 50 per cent
could be expected from a miniature generator.

1. Storage systems. The necessary 600 j of
mechanical energy can be stored in a flywheel
of reasonable dimensions and at reasonable ro-
tational speed. The basic expression for the
energy of rotation, W = Vvlw2, becomes in the
case of a simple cylinder of radius r, axial
length l, density p, and rotational frequency f,
about the axis

W = tt3 r4 lPf2.

POWER SUPPLIES

135

Thus the rotation of a steel cylinder, 1 % in. in
radius and 1 in. in axial length, represents
2,600 joules at 40,000 rpm and 2,000 joules at
35,000 rpm. The required energy could be taken
within this frequency range and a reserve con-
tent of 200 per cent would remain at frequen-
cies above 20,000 rpm. The mass of this rotor
is about 1.5 lb. The questions of adequate dy-
namic balancing and the design of bearings for
the system arise as problems in development
engineering, possibly difficult but certainly not
insoluble.

The directly coupled flywheel-alternator sys-
tem would have the advantage of permitting a
completely sealed assembly. The fuze could
carry external electric contacts by which high-
frequency alternating current could be applied
to the alternator for running it synchronously
to charge the rotor to the proper frequency of
rotation.

The system could be regarded as an a-c stor-
age battery which delivers alternating current
over the frequency range through which it has
just been charged. It has the inherent disad-
vantage of requiring an electric charging
source of adjustable frequency and reasonably
high-power output, which must be available
immediately before the launching of the fuzed
missile. If the charging system fails, the fuze
cannot be put into operation.

The operational disadvantages of field “charg-
ing” of rotors and the attendant equipment,
coupled with the possibilities of engineering
difficulties in massive fast-rotating systems pre-
cluded the serious pursuit of this method for
driving generators. It nevertheless appears
feasible, especially where the complete sealing
of a generator powered fuze is required.

Energy content calculations indicate that a
wound spring of adequate capacity would be
prohibitively large. The energy content of a
coiled clock spring is given approximately by

w BTLS2
TF = -QE~>

where B is the breadth, T the thickness, L the
length of the spring, S the applied stress, and
E Young’s modulus for the material.

Since BTL represents the solid volume of

spring material, the energy density of spring
material is given by

V 6 K

In the case of spring steel stressed nearly to
the elastic limit (S = 2 X 105 psi and E — 3 X
107 psi)

y = 25 joules /in.3 (approximately).

Assuming a 50 per cent efficiency of space uti-
lization for the spring system, 48 cu in. are re-
quired for an energy content of 600 joules. This
value is an order of magnitude too great for
warranting its consideration as an energy
source for the fuze power supply.

The use of a compressed volume of air for
energy storage appears theoretically possible
but requires extremely high pressures for hold-
ing the reservoir dimensions within permis-
sible limits. Even assuming that the release is
slow enough to approach isothermal conditions,
a reservoir of 3 cu in. capacity would have to
contain air initially at 100 atm to drive a 67 per
cent efficient turbine for an adequate energy
delivery coincident with a pressure drop to 16
per cent of the initial value, and this would
correspond to an energy reserve of only about
80 per cent. Furthermore, this system of
energy storage would introduce considerable
engineering difficulty in the design of the high-
pressure air flask, reduction valve and turbine
system. It was not given consideration in the
power supply development.

Another somewhat different storage method
has been proposed by various participants in
the program. This involves conversion of chem-
ical energy to mechanical energy. A slow-burn-
ing powder might be used to provide the driv-
ing power for a generator. Detailed considera-
tion of the method has not been made.

2. Wind-driven systems. For driving a
power supply generator it appeared most ad-
vantageous to use a vane or turbine driven by
the wind stream of the missile in flight. This
side-steps the requirement of storing within
the fuze sufficient energy for operation during
its entire service period, but instead imposes
the demand that the wind drive supply adequate

SECRET

136

ELECTRONIC CONTROL SYSTEMS

mechanical power to the generator at all times
during its service period. Thus, the spread in
air travel characteristics of various projectiles
under various applications becomes a compli-
cating factor.

A projectile moving with air velocity v and
carrying a vane or turbine of efficiency e devel-
ops power in the turbine shaft with an attend-
ant incremental force of drag f(1 on the pro-
jectile.

Thus,

P = efdv.

The deceleration a upon the projectile of mass
m is

a = U = P_

m emv

For design purposes, P has a value of 10 w, so
that an electric output of 5 w is available from
a 50 per cent efficient alternator. The aerody-
namic efficiency e of the vane or turbine will
vary with airspeed and with load, but for all the
situations of service to be met by the fuze it
should exceed 50 per cent. Using these bases,
the drag effect of the vane can be approximated
for various projectiles of minimum sizes and at
minimum airspeeds as follows.

Mass,

Airspeed
approximated
minimum
during serv-

Deceleration
due to drag

loaded

ice period

of vane or

Projectile I

(approx.)

of fuze

turbine

Mortar shell

M-43, no incre-
ment of charge

7 lb

150 fps

0.14 g

Bomb M-30, re-
lease at 150
mph 2,500 ft air
travel

100 lb

350 fps

0.004.a

Rocket T-22, at
extreme range

40 lb

500 fps

0.007#

The incremental drag due to a power supply
vane system is obviously of no importance when
compared to other sources of drag in bomb and
rocket applications. In the case of the lightest
mortar shells, it is evident that the drag may
be great enough to cause a measurable shorten-
ing of range. However, even here the effect
does not appear sufficiently serious to outweigh
the operational and constructional advantages
which the wind-driven system affords.

The system requires the design development
of vanes or turbines suitable for use with the
various bombs, rockets, and mortar shells.
Other design problems on high-speed bearings,
coupling elements, vibration isolation, etc., are
inherent to the rotary alternator rather than
the wind drive.

Selected Methods. Three methods of obtain-
ing electric power for the fuzes were selected
for intensive investigation. These were

1. Dry battery,

2. Reserve battery,

3. Wind-driven generators.

The first of these was selected for reasons of
expedience and was used in the T-5 and T-6
fuzes. The third method was used in all later
fuzes. The second method was pursued until it
was demonstrated that wind-driven generators
were practicable, at which time further work on
reserve batteries was discontinued. Summaries
of the work on these methods are given in the
next three sections.

Dry Batteries

Dry batteries were selected for use in power
supplies for early experimental fuzes and for
production fuzes T-5 and T-6. They were se-
lected because they were immediately available
in quantity. Their limitations at low tempera-
ture or in delayed service were recognized.

Development work on dry batteries fell into
two major categories: (1) the assembly and
packaging of the best available cells into a me-
chanically suitable power supply unit, and (2)
an electrochemical study of the individual cells
with the aim of improving their character-
istics.

The first battery packs were improvised as-
semblies of commercial miniature dry cells.
These were to meet the urgent need of power
units to permit proof testing of early experi-
mental electronic assemblies of the fuze. Their
design and operating characteristics were con-
ventional and warrant ho particular comment.

BA-55. One basic production dry battery
pack was developed. This was made in two
models: the BA-55, for powering rocket fuzes
T-5 in plane-to-plane use, and the BA-75 for

SECRE

POWER SUPPLIES

137

T-6 fuzes in ground-to-ground use. The two
models were identical mechanically and in bat-
tery make-up but differed in their electric arm-
ing and SD circuits, as shown in Figure 53.
The electric arming and SD characteristics
have been discussed in Section 3.3.

The BA-75 unit is shown in Figure 54. The
plastic container is 2.60 in. in diameter and

TOP TERMINALS

Figure 53. Schematic circuit diagrams for pro-
duction battery packs. Top, BA-55 used in origi-
nal T-5 fuze for plane-to-plane application. (Cf.
Figure 12, Chapter 4.) Bottom, BA-75 used in
T-6 fuze for ground-to-ground application, and
with special switch in later T-5 fuze for plane-
to-plane application.

2.31 in. in height. The pin sockets on the ends
of the container provide contact with the elec-
tronic assembly and with the arming switch-
detonator assembly. The weight of the battery
is about 10 oz.

Individual cells of the battery pack were all
of the zinc, carbon, ammonium chloride type
with manganese dioxide depolarizer. The, fila-
ment supply consisted of a parallel pair of zinc

cup cells, of miniature (No. AA) dimensions.
The plate and C-bias battery consisted of four
series stacks of cake-type cells (National Car-
bon layer-built). The individual cakes were of
rectangular cross section, 0.75 in. by 0.5 in.,
with rounded corners and were a little under
0.2 in. in thickness. The four stacks, totaling
96 cells, were arranged with the two cylindri-
cal A cells in a circular array around a cylin-
drical noninductively wound paper capacitor of
1.5 mf capacity. The capacitor served as a res-
ervoir for the detonator firing charge.

Figure 54. Battery pack, BA-75. At left is
complete unit. At right is similar unit with end
plate removed. Six stacks of layer built cells for
B and C voltage, two cylindrical A cells, and
detonator firing capacitor can be seen.

The electric characteristics of the BA-55
were as follows.

A voltage 1.50 min, open circuit, 20 C, new battery.

1.20 min, 3.3-ohm load for 30 sec, new
battery.

B voltage 138 min, open circuit, 20 C, new battery.

120 min, 8,800-ohm load for 30 sec, new
battery.

C voltage 6 min, open circuit, 20 C, new battery.

Less than 2 per cent decrease from open-
circuit voltage, 4-megohm load for 14
days, 20 C, new battery.

Temperature and Storage Properties . The
BA-55 could operate at —15 C with less than
10 per cent decrease in these voltages, and after
three months storage at 20 C could operate at
20 C or at —15 C within 10 per cent of its cor-
responding voltage output when new. The fol-
lowing table summarizes the performance of
a typical BA-55.

SECRET

138

ELECTRONIC CONTROL SYSTEMS

Delayed service performance of dry batteries
is much improved if the storage is at low tem-
perature. Three years’ storage at 9 C, two
years, at 20 C, or six months’ at 40 C causes
about the same deterioration in dry batteries.
In military field storage it could be expected

Test

Initial load

Final load

temp.

Open-circuit

voltage

(C)

voltage

(3.3-ohm load)

(30 sec)

New battery

—15

(B) 134

120

110

(A) 1.54

1.30

1.20

(B) 140

132

128

(A) 1.57

1.42

1.39

Three months’ storage at 20 C

(8,800-ohm load)

—15

(B) 133

118

108

(A) 1.51

1.29

1.20

(B) 138

128

124

(A) 1.56

1.40

1.37

that at best temperatures of 20 to 25 C would
be maintained. Thus protection against battery
deterioration could best be provided by check
on batteries immediately prior to use in the
field.

In this connection the flash current delivered
by the battery through a low (0.01 ohm) re-
sistance deadbeat ammeter was used. The
BA-55 at 20 C after three months’ storage at
20 C would give a flash current of 8 amp for
the A and 0.75 amp for the B section.

Fully adequate low-temperature service was
inherently impossible with the ammonium
chloride cells of the BA-55. Although the service
was marginally acceptable at —15 C, it became
completely impossible at about —25 C where
the electrolyte froze.

Consideration of methods for keeping the
ammonium chloride cells warm during service
indicated no practical solution. Preheating was
inconvenient and uncertain. The use of ther-
mal insulation increased bulk where it could
not be tolerated. Although it was found pos-
sible to heat the battery electrically by passing
an alternating current through it up to the
start of the service period, this was inconven-
ient and hazardous when applied to live fuzes.

Improved Dry Batteries. A dry battery
power pack with satisfactory low-temperature
characteristics appeared practical only with a
basic cell which was superior to the zinc, car-

bon, ammonium chloride cell used in the BA-55.
A study of electrolytes and electrodes was un-
dertaken by the National Carbon Company.199

For low-temperature performance, calcium
chloride was found the best of the several elec-
trolytes. Acetylene black was found superior to
conventional carbons for the positive electrode.
Synthetic manganese dioxide was found su-
perior to refined natural ore (Brazil earth) for
the depolarizer. It was also established that
several small and apparently insignificant as-
sembly details required careful control for in-
suring quality performance at low tempera-
tures.

When calcium chloride cells embodying these
improvements were tested at —40 C, after 6
months’ storage at 20 C, it was found that the
A voltage fell to 1.1 v in delivering 170 ma for
15 sec, and that the B voltage fell to about
1.14 v per cell in delivering 1.25 ma for 15 sec.
Both of these values were far short of the mini-
mum requirements from the power supply and
it was decided that the prospects of developing
a satisfactory dry cell did not warrant further
investigation.

3,4,4 Reserve Batteries199

As has been stated in Section 3.4.2, the per-
chloric acid cell had satisfactory service char-
acteristics in all respects (including low-tem-
perature properties) except for an extremely
short delayed service period. The following
table compares the perchloric acid cell,3 the
common dry cell, and the common lead cell in
a few pertinent criteria.

Perchloric

Cell

Dry Cell

Lead Cell

Open-circuit

voltage

1. 8-2.2*

1.5-1. 6

2.12

Output, amp-

hr/kg

Output, whr/kg

Freezing point,

degrees C

— 60f

—25

—65

Flash current

(miniature

1.2 amp

0.01 amp

0.37 amp

cells)

at — 50C

at — 30C

at — 40C

Internal resist-

ance (minia-

1.0 ohm

150 ohms

5.6 ohms

ture cells)

at — 50C

at— 30C

at — 40C

* Depends on acid concentration.

t Minimum freezing point concentration gives — 59 C, but this is
lowered by the solution of lead perchlorate into the electrolyte.

SECRE'

POWER SUPPLIES

139

The perchloric acid cell uses lead-lead oxide
electrodes and sustains the following reaction
during discharge:

Pb 02 + 4HC104 + Pb

-H>Pb(C104)2 + 2H20 + Pb(C104)2.
The perchloric acid cell differs significantly
from the sulphuric acid-lead cell in that the lead
perchlorate evolved on discharge is soluble in
the perchloric acid electrolyte and does not
plate out on the electrodes.

Developmental work on the perchloric acid
cell was concentrated on the design of a reserve-
type battery having the same outside dimen-
sions as the BA-55. It was recognized that such
a unit would not be usable with bomb fuzes,

Figure 55. Reserve battery pack, first experi-
mental design. This unit contained B cells only.
Section of unit is at left. Glass ampule contain-
ing electrolyte occupied central space. Cell as-
sembly, without hairpin plates, is at right.
(Photograph by National Carbon Company.)

since it required setback forces for distributing
electrolyte. However this appeared to be the
only practical means of introducing electrolyte
just prior to the service period, as necessitated
by the short permissible service delay with
perchloric acid cells.

The first experimental design of a reserve
cell, built to the same dimensions as BA-55, is
illustrated in Figure 55. This used cylindrical
plastic tubes molded in concentric circles for
housing the B cells. Nickel hairpin jumpers,
having one end plated with lead, the other
plated with lead oxide, formed the series of
electrodes. Asbestos separators were mounted

between electrodes in each cell. The acid was
contained in a centrally placed glass ampule.
This broke on setback and flooded the open ends
of the cells. Subsequent ballistic forces being

Figure 56. Reserve battery subassemblies. Top,
radial plate B cell; bottom, ampule cavity and
electrolyte distributor which was used in one
experimental model. A cell was located at the
bottom of ampule cavity. (Photograph by Na-
tional Carbon Company.)

oppositely directed to those of setback drove
the acid into the individual cells.

In proof tests of these batteries many volt-
age transients were present because of im-
proper filling of the cells.

To improve this condition the radical revi-
sion shown in Figure 56 was tried. Here the B
and C cells are formed by the many rectangu-
lar nickel plates which were molded into the
plastic base, radially arrayed about the walls of
a central plastic cup. As before, filling occurred
upon ampule breakage. The acid passed through
a fine wire mesh which removed broken glass
into the flat cup A cell and overflowed the cup
walls to fill the peripheral B and C cells. Re-
tainers in the form of perforated Vinylite
Krene, 0.008 in. thick, were used in each cell. It
is notable that the nickel cell walls were plated

secr:

140

ELECTRONIC CONTROL SYSTEMS

after being molded into the plastic base. With
the cells filled with a solution of HC104 and
PbO in water, a current of 10 ma for 8 min
plated an adequate layer of Pb on one face
and Pb02 on the other. The nickel wall thus
connects the B and C cells in series.

This model showed some improvement, but
in field tests, used with fuzes on rockets, there
were still an excessive number of malfunctions.
These were undoubtedly due to voltage tran-
sients in the B supply. Although methods were
proposed to improve the battery further, de-
velopment was discontinued because the wind-
driven generator (see Section 3.4.5) had been
proved-in as a generally satisfactory source of
electric power for fuzes.

Wind-Driven Generators

General. The wind-driven power supply con-
sisted of three principal elements.

1. The driver (windmill or turbine) .h

2. The generator.

3. The rectifier and filter.

These operated so interdependently, electrically
and mechanically, that their design was neces-
sarily evolved in close coordination.

The basic mechanical design of the entire
power supply hinged upon a choice of method
for transferring energy from the airstream
around the missile to the rotor of a generator
which was preferably located in the rear of the
electronic assemblies of the fuze. One method
was to mount a driver vane on the nose of the
fuze and couple this to the generator by means
of a central drive shaft extending through the
electronic assemblies. A second method was to
admit air through intake ports or scoops in the

h The driver for the generator on most generator-
powered fuzes was a windmill mounted externally on the
front end of the fuze. Generators on other fuzes were
driven by turbines located in air ducts farther back in
the fuze. The windmills were commonly called propellers
because of their appearance. Most of the reference re-
ports used the term propeller exclusively. The windmills
were also extensively referred to as vanes, a term intro-
duced by the Army. Both the terms, propeller and vane,
were used to specify externally mounted drivers. In-
ternally mounted drivers were referred to as “turbines.”
Occasionally the term impeller has been used (although
probably incorrectly) to include both types of drivers,
i.e., windmill and turbine.

airstream, pass it through a duct to a turbo-
generator assembly and thence to exhaust
ports.

The first method received emphasis since the
nose-mounted vane permitted the development
of a generator-powered fuze embodying much
of the basic design of the battery-powered T-5
fuze. Electronic assemblies could be revised to
give clearance for a central drive shaft of small
diameter. The inclusion of an air duct of ade-
quate cross section would have required drastic
revision. Additionally an awkward problem
arose in exhausting air from a fuze mounted in
the closed encasing can proposed for use. On
the basis of a nose-mounted vane, the following
fuzes were developed for use on bombs and
rockets: T-50-E1, T-50-E4, T-89, T-90, T-91,
T-92, T-51, T-30, and T-2004. Figure 57 shows
a sectional T-51 fuze as typical of the mechani-
cal design of fuzes of this class. The vane, bear-
ings, central coupling shaft, and metal encased
generator can be seen.

Figure 57. Fuze T-51 with section cut away.
Nose-mounted vane and central drive shaft to
generator rotor can be seen.

This discussion primarily covers the power
supply of nose-mounted vane fuzes, since these
were the ones produced in greatest quantity.
However, other types using turbine drive are
discussed in less detail.

POWER SUPPLIES

141

Bomb fuze T-82 was developed as a complete
departure from the nose-mounted vane. This
used a central air duct from the nose to a turbo-
generator at the base of the fuze. Peripheral
exhaust ports were located in the plane of the
turbine. No encasing can was required. Figure
58 shows a T-82 fuze sectioned to expose the
air duct and turbine. A similar basic design
was used in the miniature mortar fuze T-172.
Peripheral air scoop, air duct, and exhaust
ports were used in conjunction with a turbo-
generator in bomb fuze P-4. (See Figure 48 of
Chapter 4.)

Miniature mortar fuzes T-132 and T-171 and
miniature rocket fuze T-2005 avoided both the
central drive shaft and the air duct by locating
a turbogenerator at the nose of the fuze. This

Figure 58. Fuze T-82 with section cut away.
Central air duct to turbine can be seen. Gen-
erator was immediately below and directly
coupled to turbine.

was made possible by improvements in methods
of balancing the turborotor with consequently
quieter operation and by a redesign of the fuze
antenna system. A sectioned T-132 is shown in
Figure 42 of Chapter 4. Intake ports were holes
punched in the face of the nose cap. Exhaust
ports were in the side walls of the nose cap.

A design parameter of common importance
to all elements of the power supply was the fre-
quency range in which the rotor system was to

operate. Low rotational speeds relieved bearing
requirements, produced lessened centrifugal ef-
fects and vibration amplitudes, and were gen-
erally favorable mechanically. High rotational
speeds increased the electric output from a
simple generator and also simplified filtering
because of the higher electric frequency.

However, the dominant factor in selection of
the range of operating rotational frequencies
was that the speed range be removed as far as
possible from the frequency band to which the
fuze amplifier responded. Within this band the
amplifier was highly susceptible to electric
noise, whether due to generator hum or to volt-
age induced by mechanical vibration. Ampli-
fiers for various fuzes were peaked in the ap-
proximate range of 25 to 200 c, excepting the
broad-band amplifiers of transverse antenna
bomb fuzes, which had high gain up to 300 c
(cf. Section 3.2). Since operation of the power
supply was not feasible at rotational frequen-
cies below 25 c, the design was planned with
operation above 250 c (15,000 rpm) for longi-
tudinally excited fuzes and above 333 c (20,000
rpm) for the bar-type bomb fuzes.

Constancy to about ±5 per cent was required
in the voltages from the power supply during a
service period in which the airspeeds encoun-
tered by the vanes varied by as much as 3 to 1.
The requisite voltage regulation was provided
electrically by the addition of a mesh of proper
impedance to the output circuit of an a-c gen-
erator. This obviated the requirement of incor-
porating aerodynamic or mechanical devices in
the vane or turbine for regulating its rotational
speed within close limits over a wide range of
airspeeds. For all except the miniature fuzes it
was sufficient that operation stay below an
upper limit of about 40,000 rpm.

Vane and Turbine Requirements. With re-
gard to the actual supply of power it was re-
quired that a vane or turbine under its operat-
ing load should maintain rotational speed with-
in permissible limits for all airspeeds encoun-
tered between the times of arming and of fuze
function. In most cases permissible rotational
speeds ranged from 15,000 to 40,000 rpm for
bomb and rocket fuzes, corresponding to air-
speeds of 450 to 1,000 fps for bombs and 1,400
to 800 fps for rockets. Mortar fuzes permitted

SECRET

142

ELECTRONIC CONTROL SYSTEMS

higher peak values of rotational speed because
of more compact and better balanced rotor
assemblies.

However, the vane or turbine was also to
serve as an integrator of air travel and a driver
for the arming system of the fuze. This placed
the additional requirement that vanes or tur-
bines of a given type be extremely uniform in
their rotational characteristics particularly
over the range of airspeeds met during the
arming period. This meant uniformity over the
entire speed range since bombs armed at rela-
tively low airspeeds, other missiles at higher
airspeeds. Vanes or turbines for bomb fuzes
carried the additional requirement that they
develop sufficient torque to overcome the static
load of the rotating system at an airspeed of
300 fps, minimum release speed, and yet de-
velop less than this static torque at an airspeed
of 200 fps which might be encountered in an
open bomb bay.

The torques required from the vanes and tur-
bines were small. The static torques of the ro-
tating systems of the various fuzes in no case
exceeded 3 in.-oz and averaged about 1.5 in.-
oz.151 Running torques were about this latter
value.48 At the minimum operating speed of
15,000 rpm this was equivalent to a mechanical
input of 16 w, a figure consistent with the elec-
tric demand of about 7 w at an expected effi-
ciency of 50 per cent, including frictional and
other losses.

Vanes and turbines having the required op-
erating characteristics were developed for
quantity production as follows : vanes for bomb
and rocket fuzes, moldings of phenolic plastic;
vanes for bomb and rocket fuzes, punchings of
sheet steel and Duralumin; turbines for bomb
fuzes, aluminum castings and sheet steel punch-
ings ; turbines for mortar fuzes, aluminum
alloy castings.

Vane and Turbine Design. Representative
plastic vanes mounted on T-50 and T-51 fuzes
are shown in Figure 59. These had three
equally spaced blades with an effective diame-
ter of 2.5 in. The blade surfaces were helicoids
so that the vanes could be removed from a one-
piece mold with a screw motion. Vanes having
6, 9, and 12 in. of lead (i.e., helical advance in
one revolution) were used to provide the re-

quired assortment of rotational speed charac-
teristics. These are shown in the curves of Fig-
ure 60 which also includes the characteristics
of a typical metal vane.

Figure 59. Plastic vanes on bomb fuzes. Left,
T-50-E1 ; right, T-51.

The metal vane is shown in Figures 19 and
20 of Chapter 4. These were 2 in. in diameter
and carried 10 blades bent to angles of 55 or

Figure 60. Rotational speed versus air speed
for various vanes on T-50 fuze. A, plastic vane,
6-in. lead; B, Duralumin vane, 2-in. OD, 10
blades, 55-degree lead angle ; C, plastic vane, 9-in.
lead; D, plastic vane, 12-in. lead; A, C, D from
field test data of reference 27; B from field test
data of reference 28. All tests on M-81A bombs.

65 degrees relative to their original plane. The
55-degree metal vane was slightly faster than
the plastic of 9-in. lead. The 65-degree metal
was about equivalent to the plastic of 12-in.
lead. Steel or Duralumin were used for the

SECRET

POWER SUPPLIES

143

metal vanes. Brass was satisfactory in opera-
tion but was too susceptible to deformation in
handling. In some metal vanes, short radial
ribs embossed along the center line of each
blade at its narrowest section were found to
increase the rigidity and eliminate the tendency
toward blade flutter. (See Figure 19 of Chap-
ter 4.)

Both plastic and metal vanes were used on
bomb fuzes. A tabulation in detail is given in
Section 5.5 on fuze data sheets. The operation
of the vanes was affected by the airflow proper-
ties of the bomb with which they were used.
They ran slower on larger bombs. The slowing
was approximately over a 10 per cent range for
the 100- to 500-lb bombs, another 10 per cent
for the 1,000-lb and still another 10 per cent
for the 2,000-lb bomb. This effect was of limited
consequence for the ring-type fuzes which
were designed for use on particular bombs.

However, bar-type fuze T-51 was designed
for universal bomb service and used a broad-
band amplifier permitting 20,000-rpm mini-
mum vane speed during service. Here a plastic
vane of 6-in. lead was used. Operating speeds
for this are shown in Figure 61. The bomb is

VELOCITY (FT/SEC)

Figure 61. Rotational speed versus air speed
for plastic vanes on T-51 fuze. A, 6-in. lead,
M-81A bomb; B, 9-in. lead, M-57 bomb. A from
field test data of reference 35. B from field test
data of reference 30.

the M-81A, 260-lb fragmentation type. The
curve for a 9-in. lead vane, also open-mounted
on T-51, is shown for comparison. The extreme

speeds of rotation for high-altitude release were
a necessary concession to the attainment of
high rotational speed at arming for low-alti-
tude releases, particularly on larger bombs.

Figure 62. Rotational speed versus air speed
for metal vanes on T-30 fuze. A, steel vane, 2-in.
OD, 10 blades, lead angle 55 degrees; B, steel
vane, 2-in. OD, 10 blades, 65-degree lead angle.

A from field test data of reference 31. B from
field test data of reference 32.

Metal vanes were used on rocket fuzes T-30
and T-2004. The 55-degree blade angle proved
suitable for the range of airspeeds encountered.
A typical speed characteristic of the 55-degree
metal vane is shown in Figure 62. The charac-
teristic of a 65-degree blade is included for
comparison.

The T-82 turbine is shown in Figure 63. The
die-cast aluminum base is 2 in. square and
carries four fixed blades and four alternately
placed lugs to which were affixed blades of
steel clock-spring ribbon. In experimental de-
velopment light springs were used to provide
a regulating effect. At increased rotational
speeds the blades were deflected toward a radial
position both by centrifugal force and by the
increased air impact. This reduced their effi-

SECRET

144

ELECTRONIC CONTROL SYSTEMS

ciency as driving blades and also caused them
to throttle the flow from the adjacent fixed
blade. However, in the production of T-82, pos-
sible regulation features were passed over in
favor of the assurance of greater operating
uniformity attainable with heavy springs.

Figure 63. Turbine mounted on base assembly
of bomb fuze, T-82.

Nevertheless, as is shown in Chapter 5, uni-
formity of speeds for the T-82 turbine was
appreciably less than for other fuzes. A typical
speed characteristic of the T-82 turbine is
shown in Figure 64.

The turbine for mortar fuze T-132 was an
aluminum casting in the form of a circular base
1% in. in diameter, carrying eight blades
shaped as radial spirals. The speed character-
istic of the turbine for selected extremes of fir-
ing parameters is shown in Figure 65. The
curves show rotational speed against time of
flight. Extremes of airspeed are approximately
2,000 and 200 fps.

Bearings. Bearings for the rotating system
were called on for high-quality performance
under severe operating conditions even though
for a very short overall period. They were to
support the vane or turbine and generator
rotor at speeds to 40,000 rpm or faster. They
were to take axial thrusts of as much as 15 lb
from the airstream and radial thrusts of as

much as 3 lb per 0.001 in.-oz of unbalance in the
rotor at top speed. They were to introduce a
minimum of vibration and electric noise into
the electronic system.

Fuzes in final production used commercial
miniature precision ball bearing assemblies or
cushion-mounted Oilite bronze sleeve bearings
in conjunction with accurately balanced rotary
elements. Although the desirability of such
bearing systems had been apparent since early
in the program, precision ball bearings had not
been immediately available in the necessary
quantity nor had equipment suitable for rapid
production balancing operations. Pending pro-
curement of the former and development of the
latter, fuzes of the T-50 and T-51 design were
put into production, using improvised ball bear-
ings and Oilite sleeve bearings. The success
attained with these fuzes was due to the careful
attention in production to dimensional toler-
ances on components and subassemblies of the

Figure 64. Rotational speed versus air speed
for turbine of bomb fuze, T-82. Data from field
test of reference 42.

mechanical system. This is discussed in Sec-
tions 6.4 and 6.5.

Fuzes using the nose-mounted vane required
separate bearings for the vane and the genera-
tor rotor because of the 3-in. separation of
these elements and because of the assembly
problem involved. The vane bearing took the

SECRET

POWER SUPPLIES

145

thrust of the airstream and was necessarily a
ball bearing. This was located in a strong r-f
field from the oscillator and consequently was
seated in a cylindrical brass or steel sleeve
which extended downward to shield all moving
parts of the bearing and the upper end of the
coupling shaft. Vane bearing assembly may be
seen in Figure 57. (Cf. Figure 18 of Chapter 4.)

The generator shaft took no axial load. In
early production the bearings which supported
it were Oilite bronze sleeves. Final production
replaced these with precision ball bearings for
reducing end play and mutation. A ball bearing
generator is shown in Figure 70 of this chapter
and Figure 27 of Chapter 6.

Figure 65. Rotational speed versus time of
flight for turbine of mortar fuze, T-132. Curves
are for M-43C mortar shell fired at quadrant
elevations and with propellant charges as indi-
cated.

The coupling shaft between the vane and
generator transmitted a normal starting and
running torque of no more than 2 in.-oz. When
the fuze vane was freed from the block of an
arming delay mechanism after high airspeed
had been reached, the starting torque could ap-
proach 2 in. -lb. A metal shaft could not be used
because of the noise and loss it introduced in
passing through the r-f field of the oscillator.
The required strength in the permissible small
diameter was obtained by the use of rag-filled
phenolic resin and other plastics. Even with a
plastic shaft it was found necessary in the T-51
to electroplate a floating shield of copper inside
the %-in. sleeve surrounding the shaft to re-
duce the loss modulation which was introduced
at rotational frequency.

In some vane bearing designs two ball-bear-

ing assemblies were used. In this case both vane
and generator rotor spun on established axes.
The coupling shaft between their respective
shaft ends was indexed on center pins which
were fitted loosely enough to allow for the
maximum tolerable misalignment. In other
vane bearings a single ball-bearing assembly
was used. This was mounted on a coupling shaft
carrying the vane at its upper end and free
at its lower end. The axis of the coupling shaft
and vane was established when the free end
was connected to the generator shaft.

Fuzes employing turbogenerators required
only two bearings and no separate coupling
shafts. Ball-bearing assemblies were used suc-
cessfully with rotors which were not processed
for balancing. Sleeve bearings and a single-ball
thrust bearing were used with the precision-
balanced turborotor of mortar fuze T-132. This
is treated in detail in Chapter 4, which also in-
cludes a discussion of balancing methods and
equipment.

Dynamic Balancing. In fuzes having a nose-
mounted vane, unbalance in the vane was found
most serious in producing vibration and electric
noise. This was due to the location of the vane
farthest from the supporting base, and the
overhung mounting of the vane relative to its
bearings. The generator rotors were of about
the same weight as the vane (1 oz) but pro-
duced less noise because of their position near
the base of the fuze and their mounting be-
tween two bearings. Satisfactory operation of
a vane, either metal or plastic, was attained if
its unbalance after mounting were made less
than 0.001 in.-oz relative to its axis. The bal-
ancing of individual generator rotors was not
found to be necessary, provided careful control
of their dimensions were maintained.

Electric Design of Generator. Production
model generators were alternators with sta-
tionary armature windings and rotary fields.
Separate windings were used for supplying
plate and filament voltages. Rotors were small
disks of Alnico II or IV, magnetized with six
peripheral poles alternately of reversed polar-
ity.

The choice of this design rested on the fol-
lowing advantages:

1. It met the requirements of small size.

SECRET

146

ELECTRONIC CONTROL SYSTEMS

2. It required no slip rings or commutators.

3. Its use of a simple solid metallic rotor
suited it to operation at high rotational speeds.

4. It facilitated the regulation of supply volt-
ages over a wide range of rotational speeds by
electric means, since the generated emf was
directly related to rotational speed in both fre-
quency and amplitude.

5. It was well adapted to quantity production
by conventional methods.

Extremely scanty information was available
in the technical literature on design principles
for permanent magnet alternators. These had
previously been used to a limited extent as in-
dustrial tachometers, as magnetos, and as spe-
cial purpose generators.214 The attainment of
the power-to-volume ratio required for the fuze
generator depended upon the use of relatively
new magnetic materials and the extremely high
rotational speeds involved.

The feasibility of the permanent magnet
alternator was proved by exploratory investi-
gations at the National Bureau of Stand-
ards.4- 18 Early models of generator-powered
fuzes were designed by the Westinghouse Elec-
tric and Manufacturing Company, and engi-
neering design of the generator used in produc-
tion fuzes was done by the Zenith Radio Cor-
poration.19

1. Principles of operation of the alternator.
The principles of operation of the permanent
magnet alternator were easily derived for a
highly idealized case and with several simplify-
ing assumptions. The complete and rigorous
mathematical analysis, including the effects of
nonlinearity in the magnetic circuit and in the
rectifier and hot filaments which constituted
the coupled loads, was not attempted. If the
solution were actually possible, its value is not
consistent with the labor demanded. The ideal-
ized solution was fully adequate for evaluating
the parameters of generator operation and for
indicating the principles upon which a straight-
forward experimental development could be
based.

A conventionalized diagram, applicable to the
permanent magnet alternators which were used,
is shown in Figure 66. The rotor, a six-pole per-
manent magnet disk, is located centrally within
the magnetic stator which carries the arma-

ture windings. The magnetomotive force of the
magnet passes flux through the stator to link
each of the armature turns, and for each 60-
degree rotation this flux is reversed in polarity.
The resulting emf in the armature coils com-
pletes one electric cycle for each 120 degrees
of rotation.

In the analysis which follows the six-pole
alternator is considered as the equivalent of
three series-connected bipolar alternators op-
erating at three times the actual rotational fre-

Figure 66. Diagram of six-coil generator.
Rotor, stator, and armature windings are shown.
Location of magnetic poles on rotor and flux
paths are indicated.

quency. Each bipolar alternator carries an
armature winding of (N/ 3) turns, where N is
the total number of turns for the six-pole alter-
nator. Each bipolar magnet develops magneto-
motive force M equal to that developed by an
adjacent pair of poles on the six-pole rotor. The
permeance Pm for the rotor-stator flux linkage
corresponds to that for an adjacent pair of
poles in the six-pole alternator, i.e., two of the
six rotor paths in shunt, two of the six stator
paths in shunt, and two air gaps in series.

Experiment indicated that an alternator of
this type delivered an essentially sinusoidal
current 7 into a load containing only linear ele-
ments. The generating flux <f> was then sinus-
oidal. The stator saw the rotor as a sinusoidally

SECRET

POWER SUPPLIES

147

varying mmf M of internal reluctance inde-
pendent of angular position.

Thus, neglecting phase,

I = /max e**,

</> = 0max (40)

M = Mmax ejwtj

where w = 2nf (electrical) — 6jt/ (rotational).
The total flux linking an adjacent pair of arma-
ture coils with their pair of rotor poles is due to
the rotor mmf and to the current in the coils
themselves.

<t> = MPm + (Pm + P,), (41)

where Pm is the effective permeance of the mag-
netic path through magnet, air gaps, and stator
(two poles) . Pj is the leakage permeance of the
stator (two poles). N is the total armature
turns.

The emf generated in the armature is

E=~kN% (42)

where k, a proportionality factor, is equal to
10~8, and introducing equation (40),

E = —jkNw<l), (43)

combining with equation (41),

E = —jkNw | ~MPm + (P„ + p,)J. (44)

The term E drives the current / through the
internal resistance of the armature coils Rif and
the external load,

ZQ = R o + jX o,

E = I{Ri + Ro + jX 0)

combining with equation (44),

(45)

I(Ri + + jX 0)

= -jkNw | ~MPm + 0'4y— (Pm + P,)J,

j = -jkMPmNw

Ri+R0+j^^~ (Pm+P/)+X0J

| 7 | kMm ax Pm X w

| J- max | — 1 =

^Ri+Roy+^AMPw (Pm+Pi)+XoJ

(46)

In practice X0 was always capacitative. At
higher frequencies, the current approaches the
limit.

| /max |

SMr

OAN Pm + P ’
which may also be written

I t | /0max-A

(47)

(48)

From the foregoing it is apparent that with
increasing frequency the alternator becomes a
constant current source which will supply a
constant voltage to a load of fixed resistance.
The limiting maximum value of current is, by
equation (48) , directly proportional to the rotor
flux linkages with the stator turns and inversely
proportional to the total generator inductance.
By equation (47) the limiting current is in-
creased by increase in M (the rotor mmf) or
by increase in Pm (the rotor path permeance),
but is decreased by increase in N, since this
gives a square law increase in the inductance
and only a first-power increase in rotor flux
linkages.

While leakage permeance contributes no
power producing flux linkage, it is seen from
equation (46) to be as effective as rotor perme-
ance in increasing the rate at which current
constancy is approached with rising frequency.
For increasing this rate R0 and Ri should be
held to minimum permissible values.

The internal resistance Rt contains three
series components : the resistance of the stator
turns, the reflected resistance of stator and
rotor hysteresis loss, and the reflected stator
and rotor eddy-current loss. The reflected re-
sistances both increase as the square of the fre-
quency.217 It is obviously important to minimize
them, since they represent power losses which
impair rather than help regulation of load volt-
age with frequency.

In practice the design of the magnetic circuit
of the generator was worked out to provide
some excess of power at the minimum rota-
tional speed. Subsequent adjustment of coil
turns, bleeder circuit components, and the rotor
magnet strength achieved the proper output
voltages and regulation characteristics.

2. Voltage regulation. The permanent mag-
net alternators of production design were es-

1 SECRET

148

ELECTRONIC CONTROL SYSTEMS

sentially self-regulating with frequency, pro-
vided they were heavily loaded. However, the
voltage regulation was improved by the addi-
tion of capacitative reactance to the load cir-
cuit.22 This was done by either a series or a
shunt connection. The methods were used indi-
vidually or in combination.

Series connection may be evaluated from
equation (46) of the preceding section. With
X0 negative the total reactance goes to zero at
some frequency and the generator current is
limited entirely by the circuit resistance. The
value of C could be chosen to produce current
resonance just at the lowest operating fre-
quency so that transition into current, and
voltage, constancy was made more sharp. If C
were too small for the values of L and R, the
high Q of the circuit at resonance caused over-
regulation. With too large a value of C the en-
tire effect was lost.

Shunt regulation was obtained by shunting
the generator output with a mesh consisting of
a resistor and a capacitor in series. The values
were so chosen that the mesh loaded the gen-
erator only slightly to the threshold operating
frequency but loaded increasingly with increas-
ing frequency. Analysis indicated that this cir-
cuit was inherently overregulating if C were
too great and R were too small.

The discussion to this point has considered
the voltage regulation of a single supply volt-
age from a single generator winding, the high-
voltage supply. The filament winding was regu-
lated, however, by reaction from the high-volt-
age regulation since it was closely coupled to
the high-voltage winding with an iron core in
common. The coupling was not 100 per cent and
consequently the cross-regulation was not per-
fect. It was usually necessary to maintain B
voltage with slight overregulation in order to
develop adequate A regulation. Early develop-
ment model generators achieved good cross
regulation by locating the coils in the flux field
so as to produce a phase asymmetry in their
voltages.22 This was abandoned in production
models to simplify the mechanical design of the
generator.19 The C bias voltage was inherently
regulated with the B supply, since it was de-
veloped as an IR drop in the plate current cir-
cuit.

Representative A and B voltage rotational
speed regulation characteristics are shown in
Figure 67. Shunt regulation circuits for fuzes
T-50, T-51, and T-30 are shown in Figures 75,
76, and 77. Series regulation as used in fuze
T-132 is shown in Figure 78. Here the series

1.5

S £ 1.4

5 -j

m 1.3

Figure 67. Plate and filament supply voltages
versus rotational speed of generator in T-51
fuze. Dashed curve A is for filament voltage.
Solid curve B is for plate voltage. (Reference
209.)

capacitance was provided by the capacitors of
the bridge-type voltage-doubling rectifier. Com-
pound regulation, i.e., shunt and series capacit-
ance, was used in fuze T-82 as shown in the
circuit of Figure 79.

3. Rotor design. The magnetic rotor has been
considered in the case of an ideal generator as
a source of sinusoidally varying mmf having
constant internal reluctance. The rotor of an
actual generator does behave much in this man-
ner, as can be seen by reference to Figure 68,
the major hysteresis loop for the magnetic ma-
terial constituting the rotor.68 A rotor in its
matching magnetic stator (such as is shown in
Figure 66) operates on a minor hysteresis loop
4-5 while in rotation. The minor loop lies with-
in the third quadrant of the major loop specifi-
cally located according to the magnetic precon-
ditioning of the rotor. Its slope in any event is
very nearly equal to that of the major loop at
point 2.215 The rotor, for any angular position,
has an operating point at the intersection of the
minor loop with an appropriate shear line. The
shear line is a radius vector whose negative
slope equals the ratio of permeance of the space
occupied by the magnet to permeance of its ex-
ternal flux path. The end points of the minor
loop lie on the two shear lines 0-6 and 0-7,
which represent respectively maximum exter-
nal permeance (the angular position for align-
ment of rotor and stator poles) and minimum
external permeance (the mid-position of com-
plete misalignment).

ROTATIONAL SPEED (RPM*I03)

POWER SUPPLIES

149

The axis of the minor loop 4-5 must pass
through point 3, which is the intersection of the
major loop with shear line 0-3. Shear line 0-3
corresponds to the minimum external perme-
ance to which the magnet has been exposed.
Where rotors have been removed from a mag-
netizing jig and transferred openly to the gen-
erator, 0-3 corresponds to free-space permeance
and is the strongest demagnetizing force the
rotor can encounter except for the application
of a demagnetizing field from external current
turns.

Figure 68. Diagram of magnetic operating
cycle for material constituting generator rotor.
Outlying curve is major hysteresis loop for ma-
terial. Operation is on minor loop 4-5.

If the axis of the minor loop is extended to
intersect the H axis, point 8 gives the value of
virtual field intensity of the magnet. This value
multiplied by the effective length of the magnet
is the M (maximum mmf) of equation (40).
The slope of the axis of minor loop 4-5 defines
peff, the effective internal permeability of the
rotor material. Together with the effective
length and cross section of the rotor magnet
this determines the internal reluctance of the
magnet. This reluctance, the air gap reluctance,
and the stator reluctance determine Pm (rotor
path permeance for two adjacent poles) of equa-
tion (41).

For a rotor advance of 120 degrees the stator

sees one full cycle of sinusoidal mmf but the
rotor meanwhile twice traverses its unidirec-
tional loop of operation. At point 5 in the loop
the stator sees 0 mmf. At point 4 it sees a maxi-
mum mmf, either positive or negative accord-
ing to the sense of rotor-stator pole alignment.
The minor loop 4-5 defines rotor operation in
the unloaded generator. With a load applied the
rotor magnet is linked by armature current
turns and is subjected to an additional demag-
netizing force which shifts the loop down and
along its axis by an amount dependent upon
the phase and magnitude of the load current.
The use of high coercivity magnetic material,
such as an Alnico, is indicated if this effect is to
be minimized.

In the design of the generator, spatial con-
siderations dictated a maximum rotor diameter
of approximately 1 in., which, for six-pole mag-
netization, set 0.5 in. as the length of each
magnet. The minimum radial gap between
rotor and stator poles for quantity production
was set at 0.010 in., with gaps of 0.020 in. or
more considered preferable. The maximum per-
meance shear line could be roughly estimated
to have a negative slope of 25 in the case of the
0.010-in. gap, and 12.5 in the case of the 0.020-
in. gap. Reduction of the effective air-gap area
by reduction of stator pole thickness would
further reduce the slope of the shear line.

Permanent magnet steels are used most effi-
ciently at an operating point which puts their
EH product at a maximum. This corresponds
very nearly to operation with a shear line hav-
ing negative slope equal to the ratio of residual
induction to coercivity for the material.215 For
Alnico I this ratio is 16; for Alnico II, 13; for
Alnico IV, 7. All were tried as rotors during
the experimental program. Alnico IV proved
magnetically superior and was used universally
in the production generators. Cast rotors of
Alnico IV also proved mechanically stronger
than cast rotors of other Alnico types.

Salient pole rotors, like that in the develop-
ment model generator of Figure 69, were tried
as a means of increasing the effective length of
the rotor magnets. When magnetized by con-
ventional means, the internal magnetic paths
apparently jumped the tooth spaces between
poles, so that the benefit was not realized.45

150

ELECTRONIC CONTROL SYSTEMS

Since they were mechanically weaker and less
well balanced than the simple disk rotors, their
use was abandoned.

The proper selection of the A to B turns ratio
for the armature coil permitted the supply of
precisely proportioned A and B voltage
through average rectifiers to nominal loads.
The B/A ratio could be held within tolerable
limits (7 per cent) when rectifier, filter, and
load components were allowed their contingent

external coil or by the passage of alternating or
direct current through the armature winding.
By reference to Figure 68, it is seen that the
demagnetizing force moved the operating point
of the magnet along its minor loop to the inter-
section with the major loop at point 3. Further
demagnetizing force moved the operating point
down the major loop. When the demagnetizing
force was removed the magnet assumed a new
minor loop on an axis parallel to, but below, its
former axis and with operating end points on
its former shear lines.

The demagnetization produced a stabiliza-
tion of the magnet against the effects of demag-
netizing forces from momentary overload, etc.
The farther displaced its operating minor loop
from the major loop the greater was the mar-
gin of protection. Since the great percentage of
production generators required strong demag-
netization, their operation in service showed
no fatigue effects nor pole shift.

Production Models. The production model of
the basic six-coil generator, for use in nose-
mounted vane-type fuzes, is shown in Figure
70. An Alnico IV rotor in the form of a disk,
1.020 in. OD and 0.25 in. thick was used. The
stator core was a stack of five punched lamina-
tions of 26 gauge (0.0188 in.) low-silicon audio-
transformer steel, C grade. The radial air gap
between rotor and stator was 0.010 in., which

Figure 69. Early experimental generator.
Three-section stator and thick salient pole rotor
are shown.

of spread. However, the variation of magnetic
strength in individual saturated rotors, al-
though it produced no effect in the B/A ratio,
caused a spread of over 20 per cent in the com-
mon level of the voltages.08 This spread was
eliminated by designing the generator to de-
liver voltages of the required or greater value
with all saturated rotors except a small per-
centage of the weakest. After the assembly of
the entire power supply the rotors were demag-
netized individually to provide the proper oper-
ating voltages.

The demagnetizing field could be applied by

Figure 70. Production model six-coil generator
and principal components. Assembled generator
is shown at left. Rotor, mounted stator, and cover
plate are shown at right.

was maintained at +0.0030 to —0.0015 in. in
production by careful control of the dimensions
of the die cast housing and its seats for the ball
bearing assemblies. The maximum dimensions
of the housing were 2.75 in. for the diameter

POWER SUPPLIES

151

and 0.75 in. for the thickness, exclusive of shaft
extensions.

Although three coils would have been ade-
quate for intercepting all the flux of the mag-
netic circuit, small coil dimensions were pos-
sible when six were used. This permitted a re-
duced length of mean turn and a thinner gen-
erator assembly. Each coil was wound on a
plastic bobbin with a high-voltage winding of
1,940 jumbled turns of No. 41 AWG enameled
copper wire. Over this was the filament wind-
ing of 13 turns of No. 28 AWG Formvar-coated
copper wire. The six B and the six A windings
were connected respectively in series.

In operation the generator developed open-
circuit voltage in the plate winding which in-
creased linearly with rotational speed to ap-
proximately 1,000 v at 40,000 rpm. This indi-
cated the absence of appreciable core loss in the
magnetic circuit.19 The power output into rated
loads with an average saturated rotor was ap-
proximately 10 w at 15,000 rpm. This provided
an adequate margin for accepting a high per-
centage of rotors after voltage had been stand-
ardized by demagnetization.

A second production-model generator for use
in the nose-mounted vane assembly was the
single serpentine coil model shown in Figure
29 of Chapter 6. This was developed as a means
of reducing the production complexity of the
six-coil generator by the use of a single pliant
bundle wound coil which could be intertwined
about the stator pole extensions. The coil was
impregnated with varnish and baked in its final
deformed position on the stator. The flux link-
age with the serpentine coil is identical with
that of the six-coil generator, each of three
circumferential sections of the single coil being
linked by a complete flux path in the one case
and each of three pairs of series coils being
linked by a complete flux path in the other.

The rotor was an Alnico IV disk, 1.178 in.
OD and 0.25 in. thick. The stator core was a
stack of seven laminations of 0.075-in. low-sili-
con transformer steel, C grade. A rotor-stator
air gap of 0.020 in. was possible here by virtue
of the increase in rotor diameter and in thick-
ness of the stator pole face. The generator was
assembled into a case consisting of two drawn
brass cups. Maximum dimensions were ap-

proximately 2.85 in. for the diameter and 0.75
in. on the axis, exclusive of shaft extensions.

The serpentine coil included 2,700 turns of
No. 39 AWG Formvar-coated copper wire for
the B winding and 21 turns of No. 28 AWG
Formvar-coated copper wire for the A winding.
The electric operating characteristics were
essentially equal to those of the six-coil gen-
erator.

Miniature mortar fuze T-132 used the stand-
ard six-pole generator with mechanical modifi-
cations for adapting the stator assembly to a
maximum 2 in. OD and for incorporating the
rotor into a single unit with the driver turbine.
This is shown in Figure 42 of Chapter 4. The
reduction in OD of the stator assembly was
effected by removal of peripheral sections of
the lamination stack which had been used for
mounting the stator. The magnetic operation of
the stator was not significantly affected. The
rotor was reduced in diameter to 1.000 in. with
a resultant increase in radial air gap to 0.018
in. The consequent reduction in generator out-
put relative to the standard six-pole model was
corrected by appropriate revision of the recti-
fier and load circuits.

In a similar way the miniature mortar fuze
T-171 adapted the serpentine coil generator to
turbogenerator use. The rotor and the stator
lamination stack were in this case identical to
those used in T-132. For forming the serpen-
tine coil on a small radius and holding axial
thickness to the permissible maximum the
winding was distributed in two coils. The re-
sulting double serpentine is shown in Figure
35 of Chapter 4 in comparison with the single-
coil model.

The generator for miniature mortar fuze
T-172 used three coils and a multisection stator
core in a novel design evolved by the Zenith
Radio Corporation. The production model was
superficially similar to the development stator
assembly1 shown in Figure 45 of Chapter 4.

The production stator was a 0.125-in. stack
of fine transformer steel laminations in the

1 This development stator shows the pole piece ring as
continuous. Magnetic isolation of the six “pole pieces”
was effected by shading the ring at the six intermediate
points with copper straps. The production model was
superior to this model both in operational characteristics
and in simplicity of construction.

152

ELECTRONIC CONTROL SYSTEMS

shape of a ring 2.00 in. OD and carrying three
radial magnetic paths with wide pole faces on
an ID of 0.820 in. Three additional radial ele-
ments with wide pole faces were preassembled
as stacks, upon which coil bobbins were molded
and the coils wound. These coil:on-core assem-
blies were riveted into the stator ring to com-
plete the stator magnetic circuit. A retainer
ring carrying six lugs was used to brace the
pole piece “ring” and index adjacent poles for
circumstantial air gaps of 0.050 in. The re-
tainer was a nonmagnetic alloy (Advance) of
high resistivity for minimizing eddy-current
loss.

The rotor was an Alnico IV disk, 0.780 in.
OD and 0.250 in. thick. This provided a rotor-
stator air gap of 0.020 in. The large stator pole
width permitted a reasonably steep shear line
for rotor magnet operation even with the small
diameter rotor. Consequently, the electric
characteristics of this generator were similar
to those of other mortar fuze generators.

Bomb fuze T-82 used a turbogenerator seated
in the base casting which threaded into the
fuze well. (See Figure 58.) This limited the
OD of the generator to 1.33 in. and its axial
length to about the same value exclusive of
shaft extensions. A satisfactory design to the
space requirements was achieved in conjunc-
tion with a single-bobbin wound coil by a
drastic revision of the magnetic circuit.

The serpentine coil generator used a mag-
netic circuit (involving three pairs of poles)
which was restricted to one plane. The winding
was deformed from the plane to thread alter-
nately over and under adjacent stator poles.
Alternately the winding could have been held
in one plane and the magnetic circuit folded
around the coil. The T-82 generator, which is
shown with its disassembled components in
Figure 71, used a magnetic circuit which folded
around the coil and brought six stator pole
pieces of alternate polarity into alignment with
those of a magnetic rotor located coaxially at
the end of the coil. Despite its unconventional
magnetic circuit design, the T-82 had an elec-
tric and magnetic operating cycle identical with
that of the serpentine or six-coil generator.

The principal components of the magnetic
stator were a cup with three extended poles, a

spider with three extended poles, and an annu-
lar magnetic core which linked the cup and
spider through the center of the coil. The cup
and spider were drawn from a 0.062-in. sheet
of 47 per cent ferronickel. The core was turned
from first quality ingot iron.

Figure 71. T-82 generator and principal com-

ponents. Assembled generator is shown at left.
Grouped at right are spider, coil, core, rotor,
center stud with upper bearing, cup with cover
plate.

The success of this generator design hinged
on the use of 47 per cent ferronickel in the
magnetic stator. It was used unannealed after
drawing and still had extremely high permea-
bility and low hysteresis losses at the low
(2,500 gauss) flux densities encountered. In
addition, its high resistivity was of extreme
importance in reducing eddy-current losses to
a negligible value in an unlaminated assembly.

The magnetic rotor was a disk of Alnico IV,
1.140 in. OD and 0.400 in. thick, magnetized
with six peripheral poles. The radial gap be-
tween rotor and stator was 0.030 in. The large
air gap gave a very low slope to the shear line
for the rotor magnet with consequent poor utili-
zation of the magnet volume. This necessitated
the use of the thick rotor. The coil bobbin car-
ried a jumble winding of 3,800 turns of No. 42
AWG enameled copper wire for the high-volt-
age supply and over this 28 layer wound turns
of No. 28 AWG Formvar copper wire for the
filament supply.

Because of its long magnetic circuit, the T-82
generator was characterized by high leakage
inductance which limited the power available
into the operating load to 7 w with a saturated
rotor at 20,000 rpm. For this reason a com-
pound regulating circuit was used to assure

POWER SUPPLIES

153

adequate operating voltages when marginally
weak magnetic rotors were used. The high leak-
age inductance also gave a large value of volt-
age-load regulation. In consequence the T-82
rotor was demagnetized to standard output
voltage after final assembly of the fuze when
the generator operated into its actual load.

A novel experimental generator was devel-
oped as a standby for use in the T-82 fuze-well
mount. This used a stator of stacked lamina-
tions and a coplanar coaxial disk rotor of small
diameter. However, the stator carried a single
distributed winding of 90 turns of No. 20 cop-
per wire. By use of a rotor 0.5 in. thick and an
air gap of 0.010 in. a power output of 13 w at
20,000 rpm was delivered at 6 v. This was fed
to an externally located miniature transformer
whose secondaries supplied the required A and
B voltages.

The turbine-driven generator of the P-4 ex-
perimental bomb fuze differed both in design
and operating principle from other genera-
tors.204 In this case both permanent magnet and
armature coils were stationary. The emf was de-
veloped by passage of pulsating flux through
the coil core under the control of a highly per-
meable salient pole rotor which while in rota-
tion served as a periodically varying reluctance
in the magnetic circuit.

The general assembly can be seen in Figure
72. The two coils carrying distributed A and B
windings were mounted on the legs of a U-
shaped lamination stack which was eccentri-
cally located relative to a laminated rotor hav-
ing seven salient poles. A yoke containing the
permanent magnets linked the base of the
U-shaped coil core to the rotor through a wide-
angle pole face which saw one pole of the rotor
for any angular position. The two poles of the
coil cores were designed for an angular separa-
tion of one and one-half teeth of the seven-pole
stator. Thus the rotor in rotation alternately
passed unidirectional flux through the two coils
at a frequency of 7 cycles per revolution and
with 180-degree phase difference. The coils
were series-connected to deliver alternating
current at 7 times rotational frequency.

With its 36-blade 45-degree blade-angle metal
turbine the generator operated at 9,000 to 30,-
000 rpm for airspeeds of 300 to 1,000 fps, so

that the generator emf was in the range of
1,050 to 3,500 c. In this range the generator
was regulated by its own inductance and no
external regulation circuit was required. Be-
cause of the use of stationary magnets it was
possible to effect normalization of output volt-
age by use of an externally adjustable magnetic

Figure 72. P-4 power supply, partly disassem-

bled. Stator and its two armature coils can be
seen at right edge of generator assembly. (Photo-
graph by Bell Telephone Laboratories.)

shunt. The voltage and power delivered by the
P-4 generator was approximately equal to that
delivered by the rotating magnet designs.

34 6 Rectifier System

A rectifier was required for converting the
alternating high-voltage output of the genera-
tor to direct current for supplying the plate
circuit of the fuze. Filaments could be operated
on alternating current directly from the gen-
erator A winding as discussed in Sections 3.1
and 3.2. The B-supply rectifier was necessarily
of an electronic type, since the intermittent con-
tacting of mechanical rectifiers was an intoler-
able source of r-f disturbance. Thermionic di-
odes and blocking layer cells, both selenium
and copper oxide types, were considered as
rectifiers.

Full-wave rectification was a virtual neces-
sity in order to minimize ripple and obtain a
satisfactory voltage conversion from such a
high-impedance source as the generator B
winding. Three types of full-wave connection
were possible : full wave with two diodes work-

SECRET

154

ELECTRONIC CONTROL SYSTEMS

ing from a center-tapped supply, two diodes in
a bridge or cascade voltage doubler, four
diodes in a full-wave bridge.

Vacuum-Tube Rectifiers. No existing tubes
of the subminiature class were readily suited
to this service. Triodes NR3 and NS3 had ade-
quate current capacity and inverse voltage
characteristics.62 However, they were required
in multiple, and separate electrically isolated
filament supplies were required because the
cathodes were directly heated.

Blocking Layer Rectifiers (Selenium and
Copper Oxide). Selenium or copper oxide cells
rectify by preferential conduction in one sense
of direction across an interface, between sele-
nium and iron in the one type of cell and be-
tween copper and copper oxide in the other.
Having no filaments, assemblies of these cells
could be used in any of the full-wave circuits
without need for isolated filament supplies. Be-
cause of this simplicity of application and the
indications that a selenium rectifier cell of
acceptable characteristics was feasible,216 ex-
tensive effort was directed toward the develop-
ment of such a rectifier and the effectuation of
facilities for its production in quantity. This
program, which was carried on with close co-
ordination between the National Bureau of
Standards and the several manufacturers, was
largely a program of production engineering
and quality control. It is discussed in more de-
tail in Chapter 6.

The selenium cell shown in Figure 29 of
Chapter 6 was developed as the basic component
for all the power supply rectifiers. An assembly
of 20 cells in a full-wave bridge is also shown.
Other production assemblies were a 24-cell full-
wave bridge and a center-tapped 20-cell stack
for use as a bridge doubler. The individual cells
were 0.28 in. in diameter and 0.030 in. thick.
The effective rectifying zone was a central circu-
lar area of 0.075 sq in. A typical static voltage-
current characteristic for the selenium cell is
shown in Figure 73. For comparison the static
characteristics of a type-AQ copper oxide cell
of similar dimensions is included.

Although the cells had nonlinear character-
istics for both forward and reverse current,
values of effective forward or reverse resist-
ance could be defined for any static operating

point as the simple ratio of E to I. Similarly a
value of effective resistance could be ascribed
to a group of several series-connected cells for
specified operating conditions. A full-wave
bridge (cf. Figure 75) comprising four sym-
metrical arms, each arm of forward resistance
Rf and reverse resistance Rb, could be approxi-
mately represented62 by a full-wave bridge con-
taining four ideal rectifiers, having a resistance
2 Rf in series with the load and a resistance
(Rb/2) in parallel with the source.

This clearly shows the significance of both
forward and reverse characteristics of the cells
in determining the output voltages, particularly
when a high source impedance is used. In the
fuze power supply the B winding was a high-
impedance source tightly coupled to the A sup-
ply. Here a decrease in the effective Rb of the
bridge loaded the B winding more heavily and
by reflection loaded the A winding. Conse-

Figure 73. Static characteristics of blocking
layer rectifier cells, 7 mm in diameter. Curve Se
is for selenium cell. Curve AQ is for type AQ
copper oxide cell. Data from reference 64.

quently both A and B voltages decreased. An
increase in Rf decreased the B voltage but light-
ened the load on generator so that the A voltage
increased.

In the development study which was prima-
rily the statistical analysis of extensive experi-
mental data, measurements were made on the
static characteristics of individual cells. Recti-
fiers assembled from cells of known character-
istics were then studied in dynamic service, i.e.,
operating with typical generators under typical

SECRET

POWER SUPPLIES

155

loads. The resulting correlation65 between static
cell characteristics, observed under selected ref-
erence conditions, and operating power supply
output voltages is shown graphically in Fig-
ure 74. The limits within which individual cells
were classified for acceptability are shown by
dashed lines in the figure.15

Subsequently a dynamic test on 60 c alter-
nating current was evolved68 with proper limits
for the assembled rectifier bridges. It was found
that statistical assurance of an acceptable
bridge resulted from a distribution of 75 per
cent or more of the individual cells in Class A,
95 per cent or more in Classes A, B, and C, with
no open cells in the remainder.

selenium cells output voltages could be main-
tained within tolerable limits15 over the re-
quired operating range of —40 to +60 C.192
Copper oxide cells were not suited to use in the
power supply because both forward and reverse
resistances showed a marked inverse variation
with temperature. This resulted in the develop-
ment of excessive A voltage at extreme low
temperature.64

3,4,7 Filter and Detonator Firing System

The term filter system was used broadly to
include the several elements of resistance and

0 1 2 3 4 5 6

VOLTS FORWARD DROP PER CELL AT 20MA

Figure 74. Power supply output voltages as function of static characteristics of selenium cells in the
rectifier. Curves A refer to indicated values of filament voltage. Curves B refer to indicated values of
plate voltage. Dashed lines indicate acceptance limits for cells of Classes A, B, and C. The diagram is
from reference 68.

Selenium cells had appreciable temperature
dependence. The reverse resistance was a maxi-
mum in the 20 to 30 C range and decreased for
higher or for lower temperature.206 The forward
resistance varied in approximately inverse re-
lation to the temperature. Reverse resistance
decrease, particularly at lower temperatures,
was of primary importance in affecting the
output voltages from the power supply.69 With

capacitance interposed between the rectifier
and the power-supply output terminals. The
system served the following purposes: reduc-
tion of ripple voltage in the plate supply, de-
velopment of the required C bias potentials,
storage of charge for firing the detonator, pro-
vision for adjustment of output voltage by the
insertion of selected resistors, and provision of
an electric arming delay where required.

% SECRET

156

ELECTRONIC CONTROL SYSTEMS

The power supply filters in all cases em-
ployed an input capacitor followed by a single
L section of series resistance and shunt capaci-
tance. Since the input capacitor was fed recti-
fied pulses from the high-impedance B winding
of the generator through a relatively high-re-
sistance rectifier, it contributed materially to
ripple reduction. As a class the filters operated
with about 1 v of ripple across the input ca-
pacitor and 100 mv of ripple across the B volt-
age output at 1,500 c. This ripple frequency
corresponded to a generator speed of 15,000
rpm, the minimum required for operation. At
higher frequencies the ripple attenuation was
proportionately greater.

The C bias potentials for thyratron and am-
plifier were developed across a resistor in the
negative leg of the power supply filter. A single
voltage of approximately —7.5 v was adequate
for those fuzes having high-resistance grid cir-
cuits in their amplifiers. Here the full-bias
voltage was applied to the thyratron grid and
a high-resistance voltage divider contained in
the amplifier supplied approximately —1.5 v to
the amplifier grid. In fuzes T-51 and T-82 which
used low-resistance amplifier grid circuits, sep-
arate resistors in the negative leg of the filter
provided the two bias voltages.

Due to the steep load-regulation character-
istics of the power supply an undesirable spread
in output voltages would result from the nor-
mally encountered spread in the plate current
demands of the tube complements. In produc-
tion, bleeder resistors were selected to bring
each power supply to very nearly a standard
load condition. This is discussed in detail in
Chapter 6.

The requirements upon the output filter ca-
pacitor for service in detonator firing and in
delayed arming have been treated in Section 3.3.

Power Supply Circuits

Representative circuits for several power
supplies are shown in Figures 75, 76, 77, 78,
and 79. Values for the circuit components are
included.

The basic power supply was the T-50 model
of Figure 75. A typical T-50 assembly is pic-

tured in Figure 26 of Chapter 6. The limits for
acceptable operating characteristics are sum-
marized in NDRC specification for power sup-
ply PS-1.17

Figure 75. Schematic diagram of power supply
of T-50 fuze.

The T-51 power supply of Figure 76 differed
from the T-50 chiefly in respect to the C-bias
circuit. Here a low-resistance source of ampli-
fier bias was additionally provided.

Figure 76. Schematic diagram of power supply
of T-51 fuze.

The T-30 (T-2004) power supply is shown in
Figure 77. This differed from the T-50 supply
in that the detonator firing capacitor C20 was

COORDINATION OF ELECTRIC DESIGN

157

charged through a resistor R27 to provide an
electric arming delay.

The T-132 (T-171, T-172) power supply is
shown in Figure 78. This used a bridge-type
rectifier doubler. The doubled voltage was de-
veloped across capacitors C23 and C24, as

B+ G C- A- A+ TP

Figure 77. Schematic diagram of power supply
of T-30 fuze.

shown. Voltage adjustment was effected by
selection of series dropping resistor R35 and
bleeder resistor R29. Electric arming delay was
provided by the use of resistors R27 and R28
in connection with detonator firing capacitor
C20.

The T-82 power supply is shown in Figure 79.
This featured the use of compound regulation
provided by C15, C17, and R19. A low-resist-
ance bias supply for both thyratron and ampli-
fier grids was provided by resistors R7 and R14.

ments involved many considerations which
generally required compromise to make them
mutually compatible. The one overall consider-
ation was the military requirement for per-
formance: the fuze must detonate a particular
missile or group of missiles when the fuzed
rounds were in a specified space region with
respect to the target. Along with this prime
requirement were generally a host of second-
ary requirements, such as (1) the fuze must
occupy a predetermined position on the round
and must fit a fuze well whose dimensions are
fixed; (2) the space allotted to the electronic
components was determined by mechanical con-
siderations which govern shape and volume;
(3) conditions of use which include temper-
ature, humidity, high-altitude operation, and
storage life; and (4) special electrical features

Figure 78. Schematic diagram of power supply
of T-132 fuze. This is the same as that for fuzes
T-171 and T-172.

3 5 COORDINATION OF ELECTRIC DESIGN^

3,5,1 The Coordination Problem

The design requirements for the various sub-
assemblies of radio proximity fuze were deter-
mined on the basis of somewhat arbitrary deci-
sions concerning expected performance for each
part of the fuze. Coordination of these require-

relating to time of activation, circuit switching
for arming or self-destruction, and simultane-
ous use of a quantity of fuzed rounds without
mutual interference. Mechanical design is dis-
cussed in Chapter 4 ; it will be referred to here
only with respect to the limitations which me-
chanical problems imposed on electric design.

A logical method of discussing design coor-
dination is to separate the fuzes according to
intended application (antiaircraft or ground
approach) and also according to the type of
missile on which they were to be used (bombs,
rockets, or mortars). However, in the actual

j This section was prepared by the editor with the aid
of W. S. Hinman, Jr., Chief Engineer of the Ordnance
Development Division, National Bureau of Standards.

SECRET

158

ELECTRONIC CONTROL SYSTEMS

course of development during World War II,
each fuze, regardless of its application, was
built largely on experience and -designs which
accrued from previous work. The factors in-
fluencing design depended so much on the state
of the art that for the following discussion a
chronological order is preferable. The presen-
tation will be simplified by confining the discus-
sion to the major projects.

Since one of the objectives of this section is
to tie together the preceding four sections of
the chapter, frequent references to these sec-

B+ PENTOOE C- A+ TP

Figure 79. Schematic diagram of power supply
of T-82 fuze.

tions and other chapters in the volume will be
necessary. The reasons for concentrating on
doppler-type radio fuzes have been discussed in
Section 1.2, and those reasons will not be re-
peated here.

Battery-Powered Rocket Fuze

The T-5 fuze for the M-8 rocket was devel-
oped for antiaircraft use. The requirement (see
Section 1.1) was that the fuze detonate the
rocket in the vicinity of an aircraft target
and within the lethal range of the rocket’s
fragments. Although the rocket’s fragmenta-
tion pattern was unknown, it was assumed on
the basis of anticipated performance of the
rocket that the region of greatest fragmenta-

tion density would be between 60 and 70 de-
grees off the forward axis of the rocket (see
Section 1.3). It was next decided on the basis
of knowledge of the radiation patterns of lin-
ear antennas and of experimental investigations
(see Section 2.8) that by using the rocket as
an antenna, proper directional sensitivity could
be obtained.

Size and Location. Various methods of ex-
citing the rocket as an antenna were investi-
gated, but it was readily appreciated that the
optimum location of the fuze from mechanical
and service viewpoints was the nose of the
rocket. Hence methods of exciting the rocket
from the end were developed (see Section 2.7)
which would give proper loading for the oscil-
lator and the desired directional sensitivity.
When it was established that a nose location
for the fuze was compatible with electric de-
sign, dimensions for the fuze were worked out.
Since the rocket was also under development at
the same time, it was relatively easy to coor-
dinate the fuze requirements with rocket de-
sign. However, space limitations were very im-
portant, since the larger the fuze, the less high
explosive the rocket warhead would carry.
Enough exploratory work on circuits had been
done at Division 4’s central laboratory at the
National Bureau of Standards with hearing-
aid type tubes to establish minimum space re-
quirements for the working part of the fuze.
Also, the National Carbon Company had devel-
oped a small dry battery in connection with
Section T’s shell fuze program which was suit-
able for use as a power supply. It was agreed
that the T-5 fuze would occupy the following
volume: (1) a cylinder 2% in. in diameter and
514 in. long, interior to the warhead, plus (2)
a cone of the same diameter and 2% in. long
exterior to the warhead and conforming to the
contour of the ogive of the rocket. (See Figure
1 of Chapter 5.) It was essential that the
fuze have some external volume in order to
provide proper excitation of the rocket (cf.
Section 2.7) .

Choice of R-F Parameters. (1) Carrier fre-
quency. The miniature triodes which had been
developed (see Section 3.1.4) worked fairly
well in simple oscillator circuits at various
frequencies up to about 200 megacycles. It was

SECRET

COORDINATION OF ELECTRIC DESIGN

159

desired to select a range of operating frequen-
cies below this value at which the missile would
be approximately resonant and also which
would give the proper directional sensitivity
(see Sections 2.7 and 2.9 for theoretical dis-
cussion and Figure 33 of Chapter 5 for radi-
ation pattern on the M-8 rocket). A range of
operating frequencies was desired in order to
increase the difficulty of jamming. It was also
realized that if a range of operating frequen-
cies was allowed, the production problem might
be simplified. (Actually, as is shown in Chap-
ter 6, it was found practicable to build oscilla-
tors for fuzes with remarkably small spreads
in carrier frequency.)

Accordingly, three design centers for carrier
frequency were selected, and named, for secu-
rity purposes, Red, Yellow, and Green (see
Glossary) . All of these were near enough to the
resonant frequency of the missile to give suit-
able loading and also to give suitable directional
sensitivity.

2. Oscillator and detector. One of the most
direct methods of fulfilling the requirement
(cf. Section 1.1) that the fuze be resistant to
countermeasures is to radiate lots of power.
Accordingly, as shown in Section 3.1.1, an os-
cillator circuit was selected which would give
stable oscillation under full power. (The radi-
ated power ranged from 100 to 200 mw.) A
sharply tuned diode detector connected to the
antenna-coupling circuit gave suitable indica-
tion of proximity to a target (Section 3.1.2).
Since the fuze was intended for use on single
missiles, the tuning of the detector introduced
no serious problems.

Fly-over tests and pole tests (see Sections
2.11 and 2.12) provided basic data on the mag-
nitude of signals which could be expected in
the detector circuit due to approach to a target.
In order to trigger a thyratron at distances of
50 to 100 ft from an aircraft target, it was
evident that appreciable amplification of the
signal was necessary.

Designation of Amplifier Requirements. The
amplifier characteristics selected were such that
a single amplifier design could be used with all
three of the carrier frequencies selected.

The factors involved in designing the ampli-
fier characteristic to assist in control of the

burst surface have been discussed in Section 3.2.
As shown there, the gain-frequency curve of
the amplifier was also shaped to reject certain
undesirable signals such as vacuum-tube micro-
phonics. The requirements for overall gain were
determined by the magnitude of the input sig-
nal and of the output required for reliable op-
eration of the thyratron. As indicated in Sec-
tion 3.3, the spread in critical bias voltage for
thyratrons averaged about 0.4 v. To insure reli-
able operation, a firing signal of about ten times
the average spread value was desired. Accord-
ingly, a bias was selected so that a firing signal
of approximately 4 v was required. Under these
conditions, a single-stage amplifier was able to
provide sufficient amplification to cause opera-
tion of the fuze at distances between 50 and
100 ft from an aircraft target.

Power Supply. The urgency of the request
for an antiaircraft rocket fuze was such that
there was no time to develop an ideal power
supply. Accordingly the small dry battery de-
veloped by the National Carbon Company was
adopted for the T-5 fuze, although its limita-
tions with regard to low-temperature operation
and shelf life (see Section 3.4.3) were fully
appreciated. Coordination of vacuum tube and
battery design yielded a 1.5-v A supply and
135-v B supply. The tubes and circuits were
further designed so that satisfactory operation
would continue as the A and B voltages dropped
to values of about 1.1 and 100, respectively.
This provision extended considerably the use-
ful range of the batteries.

It was realized that the high internal imped-
ance of the miniature B battery might make
the firing of the detonator through the thyra-
tron a marginal proposition so a detonator
firing capacitor was added to the power supply
(see Section 3.3.3). With this arrangement,
firing of the detonator was certain as long as
the B voltage did not drop below 100 v.

Since dry batteries deteriorate in storage,
the fuze was designed to allow testing of the
batteries (and also the fuze) prior to assembly
in the field (see Section 7.7).

Arming. Initial requirements were to have
the fuze arm about 0.4 sec after launching of
the rocket. To allow stable operation of the fuze
at arming, the tube filaments and circuit con-

SECRET

160

ELECTRONIC CONTROL SYSTEMS

stants were chosen so that all warmup tran-
sients of firing magnitude were over in about
0.2 sec. To give the circuits maximum oppor-
tunity for warmup, the arming switch was
arranged to close the filament circuits during
setback of the rocket (see Section 4.3.1). Later
the arming was delayed first to 0.7 sec and then
to about 1 sec, but the rapid warmup features
of the circuits were retained.

Mechanical Stability. Since a radio proximity
fuze functions when a signal of requisite ampli-
tude reaches the thyratron grid, it was impor-
tant to prevent the generation of spurious sig-
nals which would result in malfunction of the
fuze (see Sections 3.1.5 and 3.2.6). Although
proper amplifier design noticeably reduced
some spurious signals, it was more effective
to develop tubes and circuits which would not
generate spurious signals or respond to in-
duced vibration. The results of this develop-
ment have been covered in Sections 3.1 and 3.2.

The fuzes were subjected to intense vibra-
tion in flight due to air turbulence produced by
the missile and also due to vibration of the fin
structure of the missile. In very early experi-
mental fuzes, efforts were made to shock-mount
the fuze to prevent these vibrations from reach-
ing the tubes. This procedure proved unsatis-
factory, and in all final models the tubes and
other components were firmly embedded in po-
sition as a solid part of a single fuze assembly.
Embedding was accomplished with cements and
potting compounds (see Section 4.7) which had
the added advantage of preventing penetrat-
ing moisture from altering the electric charac-
teristics of the circuit.

There remained the problem of spurious sig-
nals generated in the missile itself. Loose fins
on the rocket could produce variable electric
contacts and consequently variations in the
impedance of the rocket antenna, which would
trigger the fuze (see Section 9.2.2). Afterburn-
ing of the rocket powder produced trails of
ionized gas behind the rocket, which trigger the
fuze (Section 9.2.2 and also 2.13). Reduction
of these difficulties was accomplished by rede-
sign of the rocket in cooperation with repre-
sentatives of the Ordnance Department and
Division 3, NDRC.

Coordination of Development Groups. Nu-

merous laboratories worked on various phases
of the T-5 development for Division 4. Their
efforts were coordinated in the division office
with the assistance of the division’s central lab-
oratory at the National Bureau of Standards.
In designing the fuze for production, one manu-
facturer handled the container for the fuze,
one the arming switch, one the battery, and five
worked on the electronic unit. Since each of the
latter had facilities which were best adapted to
certain types of construction, the need for im-
mediate production overbalanced the desire for
production uniformity, and some three differ-
ent structural designs were worked out. Each
company was allowed to use that design which
was best suited to its facilities. However, all
companies were required to hold to the same
performance specifications and to hold essen-
tially the same overall dimensions. The differ-
ence of design did not result in any material
difference in field performance; the production
of all manufacturers gave a relatively high level
of performance in proof tests (see Section
9.2.3).

3'°'3 Generator-Powered Bomb Fuzes
Ring Type

A request for an air-to-air bomb fuze was
made near the end of the T-5 program. This
application meant that a longitudinal antenna
was essential (cf. Sections 1.3 and 2.8). Prog-
ress on the development of a wind-driven gen-
erator had advanced to the stage where it ap-
peared practicable to use it for the power
supply. It appeared expedient to use the same
type of circuits and general layout which had
proven practicable on the T-5 project.

An essential difference between the require-
ment for the bomb fuzes and the T-5 fuze was
that bomb fuzes were to be used on a variety of
missiles of sizes from 100 to 10,000 lb.

After development was fairly well advanced,
the requirement was changed to an air-to-
ground application. In order to take advantage
of the work which had already been done, it
was demonstrated that the amplifier alone could
be redesigned to give acceptable air-to-ground
performance. As a parallel but lower-priority

COORDINATION OF ELECTRIC DESIGN

161

project, work was started on a transverse an-
tenna (bar-type) fuze (Section 3.5.4). The fol-
lowing discussion refers only to the air-to-
ground application.

Size and Location. The bomb fuzes were in-
tended for use on existing missiles so the fuze
was dimensioned to fit into standard fuze wells.
Since most bombs were designed to carry nose
and tail fuzes, there was a choice as to location
for the proximity fuze. As shown in Figures 21
to 24, Chapter 2, the radio sensitivity with an
end-fed antenna is greatest away from the ex-
citing end. Therefore, a tail location for the
fuze would give greater sensitivity. However,
proximity of the fuze and its antenna to the
fin structure, which was known to vibrate in-
tensively during flight, led to the conviction
that such a location would produce malfunc-
tioning of the fuze.

Another consideration was that of the length
of the fuze beyond the bomb. The fuze antenna
must be spaced and insulated external to the
missile in order to properly excite it as an an-
tenna (see Section 2.7). Because of the pos-
sible shielding effect of the fins (see diagrams
in Figure 16, Chapter 2), a greater extension
was required for a tail fuze than for a nose
fuze. Furthermore, the required extension
would make the overall length of a tail fuze
(since it would be anchored to the bomb in the
rear fuze well) several times greater than a
nose fuze. The great length would make it
much more susceptible to vibration.

These considerations led to selection of the
nose location for the most intensive develop-
ment. Nose-mounted bomb fuzes with longitu-
dinal antennas were generally referred to as
T-50 type or ring type (see Figure 5, Chap-
ter 1).

Dimensions of the nose-mounted fuze exter-
nal to the fuze well were fixed as follows. A sur-
vey of clearances in the bomb bays of various
bomber aircraft led to the conclusion that ex-
tensions of more than 5 in. beyond the nose of
the bomb would lead to difficulties in stowing
fuzed bombs. Accordingly, the length of the
fuze external to the bomb was required to be
less than 5 in. The external radial dimension
was relatively unimportant. However, a di-
ameter of 3 1/2 in. (approximately) was found

adequate to hold the fuze and was adopted as
standard.

Ballistic tests showed that the size and shape
adopted for the fuze did not appreciably change
a bomb’s flight. Thus, the VT-fuzed bombs
could be used with standard bombing tables.

Some work on lower priority was done on tail
fuzes. There was a requirement for air-burst
fuzes for large blast bombs (4,000- and 10,000-
lb) . For this application there was difficulty in
exciting the missile with a nose fuze and it was
planned to build a special antenna system, as
part of the fuze, in the large tail structure.
Considerable work was done, but the project
(fuzes T-40 and T-43)199 was curtailed on the
basis of incomplete reports that there was no
advantage in air-bursting blast bombs (see
Section 9.4.5). When the advantages of air-
burst blast bombs were finally established, the
T-51 fuze development was well enough ad-
vanced for considered use on the big bombs.

Another tail fuze project was for a 90-lb
fragmentation bomb. In this application it was
planned to use a special nonconducting fin on
the bomb. Details of the work which was still in
progress at the end of World War II are given
in reference 196 of Chapter 2. One major ad-
vantage of a tail-mounted fuze on an air-burst
fragmentation bomb is the increased lethality
of the weapon. In most bombs, the greatest den-
sity of fragments is away from the point of
detonation (cf. Figure IB, Chapter 1) and nose
initiation of the explosion is therefore desir-
able for air-burst bombs. Various schemes were
tried for obtaining tail initiation for bombs
when used with the nose-mounted proximity
fuzes.

Choice of R-F Parameters. (1) Carrier fre-
quency. The requirement that the fuze operate
on more than one bomb presented a problem
in the selection of an oscillator frequency. As
has been shown in Chapter 2, both the direc-
tivity pattern and the radiation resistance
change appreciably with bomb size. There was
no singly practicable frequency (at the time)
which would be satisfactory on all the bombs.
A very low frequency (wavelength long com-
pared to the bomb’s length) would give reason-
ably uniform performance on a variety of
bombs, but the radiation resistance would be

162

ELECTRONIC CONTROL SYSTEMS

intolerably high. Circuit techniques and r-f in-
sulating materials available at the time led to
the conclusion that radiation resistances in ex-
cess of 30,000 ohms would be impracticable,
due to loss in sensitivity.

The compromise solution was the selection of
two frequencies, one for the 500- and 1,000-lb
bombs and one for 100- and 260-lb fragmenta-
tion bombs and the 2,000-lb bombs. The fact
that the latter bomb was about twice the length
of the 100- and 260-lb bombs made a single fre-
quency for those bombs practicable. (See Figure
16 of Chapter 2 for drawings of bombs.) The
frequencies selected were designated as White
(see Glossary in Appendix 1) for the first appli-
cation above and Brown for the second. The first
production models of the fuzes were designated
as T-50-E4 and T-50-E1, respectively. Although
the details of the argument leading to the selec-
tion of these frequencies are too lengthy to give
here (see reference 8 of Chapter 2) the basic
data on which the argument was based are in-
cluded in figures in Sections 2.7 and 2.8.

Toward the end of World War II, circuit de-
velopment had advanced to the stage where ade-
quate r-f sensitivity could be obtained at higher
radiation resistances.118 It was then shown that
a single frequency designated as “Brown minus
20” would be practicable for the bomb sizes
100- to 2,000-lb, inclusive.140

2. Choice of circuit. The oscillator-diode cir-
cuit used in the T-5 fuzes was selected for use
in the first T-50 fuzes. Tuning of the diode
circuit presented some problem, since each fuze
was intended for use on more than one missile
and tuning could be optimum on only one. The
methods of resolving the tuning compromise
are discussed in reference 31 of Chapter 2 and
the selected procedures for tuning are listed in
Section 7.2.

3. Antenna design. The evolution of the an-
tenna cap for T-50 type fuzes is discussed in
Chapter 4 from the mechanical point of view.
Electrically it was desired to have a large cap
to reduce radiation resistance (see Section 2.7).
The forward extension of the antenna was
limited by overall length consideration and the
rearward extension by undesirable shunting
capacitance on the radiating load. Another
factor was the presence of the rotating wind-

mill; it was desirable that it be located in a
near-zero radio field. Compromises between the
various factors led to a ring shape for the an-
tenna of about 1-in. length and big enough in
diameter to enclose the windmill (see Figures
16 and 18 of Chapter 4). The latter figure
shows the antenna in an earlier and less satis-
factory form.

Amplifier Requirements. As shown in detail
in Section 3.3, the gain-frequency characteris-
tic of the amplifier was adjusted to compensate
for the variations in r-f sensitivity for various
terminal ballistic conditions. Here, too, a com-
promise characteristic was required because of
different r-f properties of the missiles.

The use of a generator power supply intro-
duced additional requirements on the ampli-
fier :

1. A very sharp reduction in gain above the
pass band was necessary in order to reduce the
response to hum and ripple from the generator.
In addition, hum injection circuits were em-
ployed to reduce the net effect of hum at the
thyratron grid. These were incorporated in the
feedback network of the amplifier.

2. The generator power supply made it pos-
sible to obtain fuze operation at very low tem-
peratures (—40 degrees) as was desired by the
Services. Accordingly, the components of the
amplifier had to be selected with due regard to
their temperature coefficients in order that the
amplifier would perform properly over a wide
range of temperatures.

3. Although voltage regulation circuits were
employed as part of the power supply, they
were not perfect and variation in supply volt-
age was inevitable. Thus, the amplifier design
had to be arranged so that the essential gain
characteristics would persist over a range of
supply-voltage variations.

4. An average effective holding bias of about
4 v was selected for the thyratron as for the
T-5 fuze. However, considerations leading to
this selection were appreciably different in the
case of generator-powered fuzes. The variable
contributions of hum and microphonics pro-
duced a range of effective critical voltages (de-
fined in Section 3.3), of about 1 v. This was
appreciably larger than the range of critical
biases for the T-5 fuze. Also, the method for

SECRET

COORDINATION OF ELECTRIC DESIGN

163

obtaining C biasing voltages yielded a spread
of bias values of about 1 v. Thus the average
effective holding bias was only about twice the
range of variations. Although a larger margin
might have been desirable, the requirements
for sensitivity were such that the margin was
made as small as was compatible with good field
performance.

Power Supply. The various design considera-
tions leading to the development of a wind-
driven generator for the power supply have
been adequately covered in Section 3.4.5. Fac-
tors relating to certain compromises in design
are as follows:

1. A supply. The supply for the tube fila-
ments was raw alternating current at 1.4 v.
Rectifiers or commutators to supply direct cur-
rent would have been unduly complicated and
it was simpler to design the circuit to operate
with alternating current on the filaments.

2. B supply. Plate power was rectified and
filtered and supplied at about 140-v average
value. Rectification and filtering were essential
for the proper operation of the types of circuits
employed. Some saving in space was effected by
using the detonator firing capacitor also as a
filter capacitor.

The rectifier was the critical element in the
power supply as regards low-temperature oper-
ation of the fuze. Although it performed sat-
isfactorily down to —40 C, requirements for
still lower temperature would necessitate rede-
sign of the rectifier. Some special circuits were
investigated for operation on raw alternating
current, but none gave completely satisfactory
performance.

3. C supply. Circuits were designed to obtain
grid-bias voltage from the B supply rather than
require a separate output from the generator.
One advantage of this arrangement was in
self-compensation. Overall sensitivity tended
to remain constant as the B-supply voltage
varied.

4. Electric frequency. It was imperative that
the frequencies delivered by the power supply
be outside the amplifier pass band. Accordingly,
the number of poles in the generator and its
range of operating speeds were selected to give
a minimum frequency of about 750 c under
operation conditions.

5. Regulation. To compensate for the fact
that the wind-driven turbine for the generator
must operate over a wide range of missile
speeds (about 300 to 1,000 fps), regulation of
the output was essential in order that circuit
characteristics remain essentially constant.
Regulation circuits developed (see Section
3.4.5) kept the A and B voltages constant over
the operating speed range within about ±5 per
cent. Also, total voltage changes over the tem-
perature range —40 to +60 C were less than
10 per cent. These changes were compatible
with good performance of the oscillator and
amplifier.

6. Mechanical stability. Perhaps the most
serious problem introduced by the generator
power supply was that of vibration caused by
slight unbalance in the rotating system. This
vibration tended to produce microphonics, par-
ticularly in the triode. Solutions were sought in
two directions; nonmicrophonic tubes and cir-
cuits, and better balanced rotating systems.
The work done on the two aspects of the prob-
lem is covered in Section 3.1.4 and in Chapter 4,
respectively. No hard and fast rules or division
of responsibility could be set for the two prob-
lems; they had to go hand in hand. The tubes
had to be good enough microphonically to oper-
ate reliably under vibration from the generator,
and the rotating system had to be sufficiently
well balanced that it would not produce micro-
phonics in the tubes. There was some indica-
tion that as balancing techniques improved, the
generator vibration became small or negligible
compared to that produced by the bomb in
flight.

The vibration problem resulted in one gen-
eral design criterion, namely, the rotational
frequency of the generator should be outside
the amplifier pass band. This was a more seri-
ous limitation on the selected range of operat-
ing speeds than the one mentioned above con-
cerning the electric frequency of the generator
output. (Electric frequency was usually three
times rotational frequency.) An upper limit on
rotational speeds was set by the durability of
the bearings and the centrifugal strength of the
Alnico rotors. In some later fuze designs, nota-
bly T-51, rotational frequency was on the upper
edge of the amplifier pass band at arming.

SECRET

164

ELECTRONIC CONTROL SYSTEMS

Arming. The arming problems in generator-
powered fuzes were largely mechanical and are
discussed in Chapter 4. The requirements for
warmup were less critical than those solved
for T-5 fuzes. In some designs there was indi-
cation that firing pulses would be produced at
arming, i.e., when the electric detonator was
connected to the circuit. Either proper circuit
layout, by-passing or choking, was adequate to
eliminate this difficulty.

In rocket and mortar fuze applications,
added RC arming was used. As shown in Sec-
tion 3.3.6, there was an inherent spread in arm-
ing times by this method unless considerable
care was taken in selecting component values.

Overall Stability. The same standards for
rigid assembly used in T-5 fuses were extended
and carried into the design of generator-pow-
ered fuzes. The problems were, of course, more
difficult because of the high rotational speed of
the power supply system. The special layout
chosen for the fuze was probably the most dif-
ficult from a stability standpoint, but it had
other advantages, as follows :

1. Circuit arrangements, previously devel-
oped, were used directly with only a minor
modification to allow the generator drive shaft
to pass down through the axis. However, it
proved desirable to shield the drive shaft when
it passed through the amplifier and oscillator
block.

2. With the windmill on the nose of the fuze,
the aerodynamic problems were simplified.
Later models located the entire power supply
in the front end of the fuze, dispensing with
the long, high-speed drive shaft (T-132) ;
others located the generator and turbine at the
base of the fuze using a central air duct for
directing air to the turbine (T-82, T-172). The
latter arrangement introduced difficulties in
circuit layout and space requirements due to
the central air duct.

Difficulties experienced with T-5 and the
extra vibration with the new power supply led
to the elimination of all plug-in connections on
T-50 type fuzes. The electric connections be-
tween the various subassemblies were soldered,
and during the various laboratory tests, sol-
dered connecting leads were used. Although this
increased the labor involved in making tests, it

insured that vibrating electric contacts within
the fuze would not introduce spurious signals.

Spurious noise signals from the missile were
another matter for consideration. Special wash-
ers were used to insure both good electric and
mechanically stable contact between the fuze
and bomb (see Chapter 4). Service instructions
advised that both the fuze and fin be firmly
secured to the bomb. Reasonably careful
wrench-tightening usually proved adequate for
good fuze performance. However, late in World
War II a new washer was produced which gave
excellent results with just hand-tightening of
the fuze (see Section 9.4.3). Occasional diffi-
culty was encountered with bomb fins (particu-
larly T-92 on M-64 bombs) which could be
eliminated only by unusual precautions (see
Section 9.4.3). In such cases, alternative fuze
designs were sought (see Section 1.5) since
redesign of the bomb’s fin structure was be-
lieved impracticable.

Coordination of Development Group. In gen-
eral, the arrangements for coordinating devel-
opment and experimental production followed
the same procedure as for the T-5. Standardiza-
tion was insisted on only when necessary, and
considerable individuality was allowed in de-
sign detail in order to make maximum use of
available facilities. Various types of oscillator,
amplifier, and generator construction are de-
scribed in Chapter 6.

3 5 4 Generator-Powered Bomb Fuze,

Bar Type (T-51)

The T-51 bar type bomb fuze was developed
specifically for air-to-ground application. A
transversely excited antenna was part of the
fuze and led to the name bar type. An under-
lying design consideration was to make maxi-
mum possible use of T-50 mechanical parts in
order to expedite both development and pro-
duction problems. Accordingly, the T-51 fuze
was mechanically identical to T-50 fuzes except
that a nose piece with transverse bars attached
replaced the ring-carrying nose piece.

Size and Shape of Antenna. The overall
length of the antenna was limited by the dimen-
sions of the smallest bomb on which the fuze

COORDINATION OF ELECTRIC DESIGN

165

was to be used. A maximum 10-in. tip-to-tip
length was imposed. (This was approximately
the diagonal width of the fins for M-30 and
M-81 bombs.) As long a length as possible (up
to one-half wavelength) was desired in order to
reduce radiation resistance and increase sensi-
tivity.

Early experimental models of transversely
excited fuzes had used screw-in dipoles, but
considerable difficulty was encountered with vi-
bration and variable electric contact. Accord-
ingly, it was decided that greatest stability
would be obtained with dipoles molded firmly
into the nose piece. High-strength and low-loss
dielectrics were desired and details of this in-
vestigation are given in Section 4.7.

Preliminary calculation indicated that the
only suitable cross section for the antenna with
adequate rigidity would be an airfoil section.
Cylindrical cross sections would have given in-
creased drag. The design chosen gave negligible
drag in tests on 250-lb bombs. The dimensions
of the section were selected as a compromise
between rigidity and shunting capacity on the
radiating load.

Electric Parameters. The first guess in choos-
ing an oscillator frequency for T-51 fuzes was
that the higher the frequency up to a value
corresponding to a wavelength equal to twice
the antenna length (suitable tube characteris-
tics assumed) , the better would be the perform-
ance. Actually, it turned out, owing to unavoid-
able electric unbalance of the antenna, that at
higher frequencies the bombs became strongly
resonant, and longitudinal excitation masked
the transverse excitation. Consequently, an
upper frequency limit was set at which bomb
resonance would not be troublesome. A lower
frequency limit was set by the allowable radia-
tion resistance, below which circuit losses were
intolerable. Designs centered around Yellow
(see Glossary in Appendix 1) but a fairly wide
range (10 megacycles or so) was permissible.

Oscillating-detector circuits (RGD) were de-
veloped to give good sensitivity under stable
oscillating conditions (see Section 3.1). No
tuning of the circuit was necessary and the
fuze was usable on a wide variety of missile
sizes (see Section 9.4.4).

The factors leading to the selection of ampli-

fier characteristics are adequately treated in
Section 3.2.

35,5 Generator-Powered Trench-Mortar
Shell Fuzes

The last Service requirement for VT fuzes
was for the 81-mm trench mortar for ground-to-
ground use. This project required a major re-
design of the fuze, since the effect of the fuze
on ballistics of the round was one of paramount
importance. Some of the mortar rounds weigh
about 8 lb, and the addition of a 2-lb fuze would
reduce the round velocity in about the ratio of
the increased weight. A fuze such as those used
on bombs and rockets would completely over-
balance the round, and even if it were satisfac-
tory from a range standpoint, the round would
be unstable in flight. It was therefore necessary
to reduce the volume of the fuze by a factor of
about 3. An additional requirement was intro-
duced by the necessity of withstanding acceler-
ations up to 10,000#, which was about 100
times greater than that which rocket fuzes were
required to withstand. Fortunately, the re-
quirements for reduced size and increased rug-
gedness were compatible. Two types of fuzes
were engineered for production because of lack
of time to work out compromises and an opti-
mum single design. One of these used a 2V2-im
transverse loop antenna (T-172) and the other
used the missile for an antenna. Both used
essentially the same fuze circuits, the difference
being similar to that between the transversely
excited and the longitudinally excited bomb
fuzes.

These fuzes were the first which were engi-
neered without much structural design rela-
tionship to previous fuzes. They did, however,
use the same type of components, there being
only a minor modification of the generator and
the rectifier assembly (see Chapter 4). There
was, however, considerable crowding together
and the clear-cut shielded separations between
radio frequency, audio frequency, and power
supply, adhered to closely in previous designs,
were not followed. In the longitudinally ex-
cited fuzes, the antenna cap was appreciably
elongated to decrease the radiation resistance,
already high because of the short missile. In

SECRET

166

ELECTRONIC CONTROL SYSTEMS

order that the space would not be wasted, the
electric assembly, including power supply, was
located within the antenna (see Figure 42, of
Chapter 4).

The longitudinal fuzes were built to two de-
signs which were externally similar but which
use radically different internal construction.
One of these (T-132) represented the first ap-
plication of a printed circuit technique (see
Chapter 6) in which resistors and their con-
necting wires were printed directly on ceramic
plates with appropriate points of plating for
the attachment of tubes, small ceramic and
paper condensers, and for the connection to the
power supply and detonator. The aim of this
printed circuit development was threefold: (1)
reduction in size through the elimination of the
bulk of individual components such as commer-

cial resistors, (2) increase in production speed,
and (3) reduction of cost. This fuze was near-
est to production status (of the mortar fuzes)
at the close of World War II.

The second version of the longitudinally ex-
cited fuze (T-171) utilized essentially the same
circuit system but different types of compo-
nents. It was felt that the printed circuit tech-
nique was something of a gamble, since it rep-
resented not only a development of the tech-
nique but of the fuze as well. For this reason,
it was deemed necessary to engineer a similar
fuze but using standard components whose per-
formance was well established. Further details
of the mortar shell fuzes, on which some design
compromises were still in progress at the end
of World War II, are given in Sections 3.1, 3.2,
and Chapters 4 and 5.

Chapter 4

MECHANICAL DESIGN

41 GENERAL REQUIREMENTS

Introduction

IN discussing the various aspects of the me-
chanical design and construction of prox-
imity fuzes, two approaches are possible. One
is to treat the problems from an abstract point
of view and to show how the final solutions fol-
lowed inevitably the theoretical considerations
involved. The second approach is to give the
history of the mechanical development of the
fuzes described in this report and to show how
each fuze was the successor to its predecessors
and how the considerations of expediency de-
termined its details.

It has been mentioned in the introductory
chapter that time and the availability of ma-
terials and tools were a controlling factor in
most of the engineering designs. This was par-
ticularly important in the mechanical design
of proximity fuzes and consequently the history
of the development is intimately related to all
the mechanical designs. It would have been
very pleasant for the designers if, at the be-
ginning of the development of each fuze, they
were given carte blanche in respect to the de-
sign of the fuze, its components, the vehicle, and
even some control as to the method of its
launching. Unfortunately, this was not the case,
particularly in the latter part of World War II,
when the specifications of suitable projectiles
and the methods of their launching preceded
the development of the proximity fuzes.

The limitations on the availability of compo-
nents and on the interchangeability of the
proximity fuzes with other types made all other
design considerations secondary. The primary
consideration was always time. It is for these
reasons that, after some discussion of the gen-
eral mechanical requirements of proximity
fuzes, the detailed treatment of each fuze will

a This chapter, except for Section 4.7 was written by
Jacob Rabinow of the Ordnance Development Division of
the National Bureau of Standards. Section 4.7 was
written by Philip J. Franklin of the same organization.

be taken in its chronological order so that the
reasons for the details of its construction will
be more readily understood.

4,1,2 Arrangement of Main Components

Let us consider first the case of the longi-
tudinally excited radio proximity fuze. Since
the vehicle itself is part of the antenna system,
it is highly desirable for the antenna insulator
to be as near to the mid-point of the total round
as possible. By the use of special projectiles
such a condition could be closely approximated.
Very early in the program both rockets and
bombs of special construction were built and
tested; but it soon became apparent that it
would be far more desirable to build fuzes
which would fit standard missiles, and the de-
velopment of all future fuzes was conditioned
by this decision.

All of the radio fuzes which went into pro-
duction during World War II were of the one-
piece type that fitted into the fuze well at the
nose of the projectile. As a result of this, the
forward antenna in all cases was a small frac-
tion of the total length of the vehicle. In the
case of the transverse antenna fuze there were
other limitations on the antenna size. Mainly,
the combined length of the dipoles should not be
over 10 in., which was less than the maximum
diameter of the M-57 250-lb bomb and only
slightly greater than the diameter of the M-30
100-lb and M-81 260-lb bomb, and the effects of
the dipoles on the fall of the bombs should not
require modification of the bombing tables.

The placement of the photoelectric fuze in
the nose of the vehicle presented a very happy
solution, since this position was the most natu-
ral for obtaining the “forward looking” sensi-
tivity pattern desired. In all the proximity
fuzes, with the exception of the T-132 and the
T-2005 which will be treated later, the general
arrangement was as follows : The antenna and
oscillator unit, or the photocell, were at the
forward end of the assembly, followed immedi-

* SECRET

167

168

MECHANICAL DESIGN

ately by the audio amplifier. The power supply,
consisting either of a battery or the generator
with its associated rectifier and filter, was
mounted below the electronic assembly and
was, in turn, followed by the arming system
and the explosive train.

4,1,3 Rigidity

A prime qualitative mechanical requirement
in the design of the fuzes was the production of
an assembly as rigid and as quiet as possible.
Since the proximity fuzes are, in general, ex-
tremely sensitive (one is tempted to say deli-
cate) devices, the problem of microphonic noise
is perhaps the most difficult one of all to solve.
When it is remembered that these fuzes move
through the air at speeds up to 2,500 fps and
contain turbines and generators rotating at
speeds reaching 2,000 rps, a much clearer pic-
ture can be had of the difficulties involved.

Dynamic balancing of the high-speed rotat-
ing elements was utilized with great success.
Special balancing equipment, which could be
easily and cheaply manufactured was developed
as incidental to the fuze program.

4,4 Size

The second major requirement in the design
of the fuzes concerned the limitations on size.
The fuzes were to be, as far as possible, inter-
changeable with existing mechanical fuzes and
were to be adaptable to the projectile with mini-
mum effect on its ballistic properties. In the
case of the larger bombs, this was not particu-
larly difficult, since the effect of the fuze on the
flight of the bomb is relatively small. However,
the attempt to match the ballistics of the im-
pact fuze with a radio fuze for the 81-mm
mortar shell was not accomplished successfully
during World War II. The fuzes that were
built were appreciably larger than the impact
fuzes and resulted in the decrease in range of
approximately 25 per cent.

Other Requirements

There were many other more or less second-
ary requirements, such as complete safety in

handling, long shelf life, ruggedness in ship-
ment and handling, ability to withstand heat,
cold, and the humidity of the tropics, and the
adaptability to as many projectiles as possible.
There were special requirements for simple
changes of fuze characteristics in the field, such
as changes in the arming time, the optional in-
clusion of self-destruction [SD] of the adapta-
tion of the fuze to air-to-ground or to air-to-air
service. How the particular mechanical require-
ments were met will be described separately for
each fuze.

I-2 SAFETY AND ARMING

Comparison with Other Fuzes

A special note should be placed here about
the general safety and arming requirements of
the proximity fuzes. With the possible excep-
tion of a time fuze, the proximity fuzes present
the most difficult problems as far as safety and
the arming characteristics are concerned. It is
quite obvious that the very nature of a prox-
imity fuze is such that when fully armed and
energized it presents extreme hazard to any-
thing in its immediate vicinity. Accordingly,
great effort was spent in keeping the various
fuzes inactive until safely away from their
point of launching. It is well to point out here
that one of the great advantages of the gen-
erator as compared with the battery is its
greater safety. A generator-powered fuze
equipped with only an electric detonator is a
very safe device when the turbine is not turn-
ing at a high speed.

Very early in the work the Ordnance Depart-
ment specified that all proximity fuzes be
equipped with powder train interrupters so
that, if the switching and electric safeties failed
in some manner and resulted in an explosion of
the electric detonator, the main explosive would
remain unaffected. This is also the principle
followed in most of the American mechanical
fuzes. The only notable exceptions to this rule
are the American bomb tail fuzes, in which the
detonators are generally in line with the main
explosive, but the striking pin is located away
from the primer until properly armed by a very

SAFETY AND ARMING

169

rugged and simple mechanism. This emphasis
on powder train safety is not seen in fuzes of
other nations, particularly those of Germany.
Both the mechanical and electric fuzes as used
by the Germans almost invariably had the ex-
plosive elements in line and relied for safety on
electric or mechanical devices ahead of the ex-
plosives.

The main objection to an in-line detonator
is the possibility of its going off either because
of a violent mechanical shock or because of the
heat resulting from fire. The advantage of such
an arrangement is the obvious simplicity and
compactness.

4,22 Difference between Rotating and
Nonrotating Projectiles

In general, the design of a safety mechanism
for a fuze should use the cardinal principle that
a fuze must arm only when subjected to all the
forces it experiences when released against the
enemy. The more varied the nature of these ex-
periences, the simpler is the problem of making
the fuze safe for our own troops. As an ex-
ample, a fuze which is fired from a rifled gun
experiences linear and rotational accelerations
of great magnitude, large centrifugal forces on
all components, and the effects of high velocity
of travel through air ; however, a bomb dropped
from a plane experiences very little linear ac-
celeration, practically no rotational accelera-
tion, and experiences the impact of a much
slower airstream. The rockets and the mortars
are somewhere between the bomb and the ro-
tating shell. Some of the rockets revolve, others
do not. Some are accelerated at approximately
10 g, others at several hundred g. The trench
mortar shell experiences accelerations up to
6,000# with no rotation. Beside taking advan-
tage of all these “natural” conditions present
in firing of the various projectiles, other arm-
ing means may be employed.

42,3 Possible Methods of Arming
Manual Arming

Manual arming is perhaps the most common.
It generally consists of manually setting the

arming mechanism to operate after launching
or actually completely arming the fuze. There
are many obvious objections to this approach,
and it was given up early in the program of this
division and was not employed in any of the
fuzes which actually reached the production
stage. Partial manual arming, such as removal
of the safety pin in the trench-mortar fuze, was
built into the T-132, but the later developments
at the conclusion of World War II made even
this manual operation unnecessary.

It is interesting to note that in the first of
the proximity fuzes designed by the British at
the beginning of World War II, the powder
train interrupter was manually armed ; that is,
the interrupter slider was moved into position
by means of a screw driver. If the round were
not fired, someone had to remember to move
the interrupter out of line.

Use of Arming Wire

In the case of our bomb fuzes we used the
traditional method of releasing the windmill or
the turbine by means of an arming wire that
was attached to the plane. While this method
of arming is open to very serious objections
and was the only one that resulted in some acci-
dents, the considerations of standardization of
the plane equipment were such that no changes
in this procedure could be made during World
War II. It is possible theoretically to arm a
bomb fuze automatically by making use of the
fact that a fin-stabilized projectile in flight ex-
periences a deceleration only in the direction
of its longitudinal axis ; in practice this is diffi-
cult because of the low value of this accelera-
tion and because it is conceivable that a plane
may move in a path and at velocities equiva-
lent to those of a freely falling bomb.

Air-Travel Devices

To insure the arming of the bomb fuzes at a
safe distance from the launching plane, air-
travel devices were connected to the windmills
(or turbines) so that a required number of
windmill turns had to elapse before the fuze
would be armed. Although this is very simple
in principle, certain difficulties arose in prac-
tice. Different operational conditions required
different distances to arming. Dive bombing

' SECRET

170

MECHANICAL DESIGN

techniques required quick arming, while forma-
tion flying required very large arming dis-
tances. This difficulty was recognized early in
the work, but the pressure of time was such
that variable arming was not introduced into
the fuzes, and supplementary devices were em-
ployed for this purpose.

One of these devices was the T-2 arming de-
lay shown in Figure 1A. This device was fas-

off. (See Figure IB.) A cover using a similar
mechanism was also developed for the T-50
series of fuzes and is shown in Figure 2.

Effect of Air Pressure

The effect of air pressure on the nose of the
projectile was also considered as a possible
source of energy for arming rocket fuzes. One
such system was built and tested (see Figure

Figure 1. A, T-2 arming device in place. This delay installed on antenna ring prevents rotation of vanes.
Dial on delay may be set in increments of air travel up to 20,000 ft. B, T-2 delay after operation. When
arming delay operates, it is detached from ring, allowing spring-loader plunger in ring to fly out, thus
unlocking vanes.

tened to the bomb fuze and was set to come off
after any desired length of air travel from 0 to
20,000 ft.10 The regular arming mechanism of
the fuze remained inoperative through this
part of the cycle and operated in its normal
manner only after the T-2 device was thrown

3), but was abandoned in favor of “setback”
devices.

Clocks and Timing Devices

The use of clocks and timing devices in gen-
eral was given very serious consideration, espe-

SAFETY AND ARMING

171

cially since most of the ballistic tables used by
the Air Forces in the dropping of bombs are
not given in terms of air travel. Nevertheless,
clocks were not used in any of the fuzes devel-
oped up to the end of World War II because
of the serious objection to their lack of safety.
A clock driven by a prewound spring is inher-
ently a dangerous device, since once started it
goes to the completion of its arming cycle. Near
the end of World War II the thinking in connec-
tion with the clocks underwent a change and,
as will be mentioned later, the trench-mortar
fuze was being redesigned to use a 10-sec me-
chanical delay in its arming mechanism. Also,
several schemes were suggested in which the
clocks would be driven by an air turbine so that
they would not be capable of operation unless
the fuze were subjected to an airstream of high
velocity.

In the case of the rocket and mortar fuzes,
main reliance for safety was placed on the use
of setback or inertia-operated devices that were
developed to a high degree of perfection. This
resulted in fuzes that were extremely safe in
handling; in fact, of all the thousands of fuzes
built, tested, and used, there was not a single
malfunction due to the failure of such a safety
mechanism.

Acceleration Integrators

A new principle of setback or inertia arming
was evolved in Division 4’s central laboratory
at the National Bureau of Standards. It con-
sists, in essence, of the incorporation of some
form of an acceleration integrator into the
fuze arming mechanism. This mechanism can
be so designed as to preclude the possibility of
the fuze’s arming, unless it attains a desired
velocity. This is a sharp departure from the
previous practice of using setback devices that
could be triggered by intense shocks of short
duration such, for instance, as are experienced
by a shell in landing on a hard surface when
accidentally dropped from some considerable
height. In rotating shells this problem has
never been serious because centrifugal force
can be used in conjunction with the setback
devices in order to make the latter shockproof.
For mechanical fuzes for the nonrotating shells
of the mortars, the setback device is made safe

against accidental shocks by the use of a manu-
ally removable arming wire. Also, the large
values of accelerations experienced by these
shells make it relatively easy to design a fairly

Figure 2. Alternative arming delay. This de-
lay completely encloses nose of fuze, preventing
air stream from reaching vanes. When it oper-
ates, it opens and flies off fuze as shown in
picture at right.

safe mechanism even if the arming wire were
not used.

When work on rockets and fuzes for rockets

Figure 3. Arming device actuated by air pres-
sure.

was begun, however, it became apparent that
setback devices which operate at values of 10
to 200p were extremely sensitive to accidental

172

MECHANICAL DESIGN

shock. The British in one of their early setback-
operated switches employed a spring retained
weight that drove a flywheel through a series
of step-up gears (see Figure 4). While this
type of device does act as a type of accelera-
tion integrator, it presents a real danger when
subjected to a very large acceleration of short
duration. Once the flywheel is started, it tends
to drive the mechanism to completion, even
though the acceleration ceases.

acts as an escapement meshing with a flutter
bar. When subjected to an accelerating force
greater than the force of the spring, the weight
moves toward the tail of the projectile in a
series of short steps. The overall cycle, there-
fore, requires an appreciable time ; if the accel-
eration is of very short duration, either because
of an accidental shock or incorrect burning of
the propellant, the weight does not reach its
extreme rear position but stops and is moved

Figure 4. Acceleration-operated arming device. This is a British design developed for use on their
rocket fuzes.

To overcome these objections a series of
arming mechanisms was devised which can be
divided into two basic types. One is a mecha-
nism containing a weight retained in its forward
position in the projectile by a spring which de-
termines the minimum value of acceleration
necessary to operate the mechanism. The
weight is connected to a toothed wheel, which

back to the initial starting position by the
spring. A typical mechanism of this type is
illustrated in Figure 5, showing the arming
mechanism of the T-4,b T-5, and T-6 fuzes. The
same escapement which is used to retard the
weight during its initial arming cycle is also

b The T-4 photoelectric fuze is described in Division
4, Volume 3, Summary Technical Report.

SAFETY AND ARMING

173

employed later to delay the final arming of the
fuze. This will be discussed in detail under a
separate heading.

The second general class of shockproof arm-
ing mechanisms employed the action of sepa-
rate inertia elements, each of which is retained
by its own spring. Consider Figure 6, which

eration must be sustained long enough for the
weight A to reach bottom and stay there while
weight B executes its full stroke against the
force of its spring. The safety of this arrange-
ment can be shown by the following example.
Assume springs of constant force. In the case
cited it would take a minimum drop of 133 ft

Figure 5. Arming device for T-5 fuzes. This device operates by integrating acceleration.

illustrates the following: A weight A which is
free to move for 1 in. is retained in its forward
position by a 100-g spring. Mounted in close
proximity to this weight is another similar
weight B with a similar 100-g spring. It can
also move 1 in. but only after the actuation of
a mechanism by the motion of A. This mecha-
nism, shown schematically in the drawing, is
designed to keep the weight B in its initial posi-
tion until after the weight A has reached its
lowermost position. A similar mechanism trips
an arming device, when the weight B reaches
its lowermost position.

Consider what happens if the whole mecha-
nism is subjected to an extremely large accelera-
tion for a very short time, as when dropped on
a very hard surface. If the drop were made
from sufficient height, the weight A would
stretch its spring ; but the deceleration would be
completely over before the weight B was re-
leased, and the mechanism would not permit
arming. For the mechanism to arm, the decel-

onto a very special kind of surface that would
decelerate the mechanism at a uniform rate of
200gr for 8 in. before the mechanism would
permit arming. This compares to a drop of
8.5 ft for a simple single-element device, such
as weight A with its spring alone.

It is, of course, obvious that more than two
weights can be thus interlocked and the safety
multiplied accordingly. Three weights inter-
locked as above would require a minimum drop
of 300 ft under similar conditions.

A practical device using this principle is
shown in Figure 7, illustrating two unbalanced
sectors, each maintained in its position by a
75 -g spring. The flanges of the sectors are so
arranged that the left-hand element must com-
plete a 90-degree motion before the right-hand
element can start. This mechanism, a switch
designed for one of the early rockets, was
tested by dropping from 100 ft onto a large
variety of surfaces, from concrete to soft earth,
without arming. It operated very satisfactorily

174

MECHANICAL DESIGN

when fired in rockets with an acceleration of
over 100#. This particular device was not used
on a large scale because its relatively short
arming time (0.04 sec at 125#) made it danger-
ous in case of motor blowups. A similar device
was also designed for the T-132 mortar fuze.
(See Figures 8 and 42.)

Figure 6. Double-action, acceleration-integra-
tor, arming device.

Other types of acceleration integrators have
also been proposed and tested. The use of dash-
pots was tried, but because of the general diffi-

Figure 7. Photograph of double-action arming
device.

culties with the sealing of liquids and with tem-
perature effects, this type of device was not
used.

424 Self-Destruction

When projectiles are fired over friendly ter-
ritory or when, for the reasons of security, the
number of duds reaching enemy territory must
be kept to a minimum, SD is required. Two gen-
eral methods of accomplishing this were em-
ployed. One was to use an electric circuit which
detonated the fuze several seconds beyond arm-
ing. This method was described in Section 3.3.

Figure 8. Arming mechanism for T-132. This
is a double-action device.

The second method is to use a mechanical de-
vice which operated a contact accomplishing
the same result. The mechanism that was used
in a small number of T-5 switches manufac-
tured by the Globe-Union Company consisted
of a long coil spring which drove an escapement
wheel for several revolutions after the comple-
tion of arming. One end of the spring was
fastened to the frame, while the other was
attached to the escapement wheel. The spring
consisted of some fifty turns. At approximately
three turns from the fixed end there was at-
tached to the spring a small silver-plated con-
tact. When the escapement made ten revolu-
tions, this contact made only a part of a revolu-
tion, since the rotational speed of any element

1ET

MECHANICAL DESIGN OF PROXIMITY FUZES

175

of such a coil spring is proportional to its dis-
tance from the fixed end. In this way, speed re-
duction was obtained without the use of gears.
A diagrammatical illustration of the mecha-
nism is shown in Figure 9 and a photograph in
Figure 10. In the T-2005 fuze a differential
screw was employed to attain the same result.
(See Figure 47.)

4-2,5 Impact Detonation

Throughout the history of the proximity fuze
development, considerable controversy existed
as to the desirability of including a mechanical
impact detonating element in the fuze. Impact

One form of impact detonation which retains
some of the advantages of overcoming duds
consists of providing the fuze with inertia-
operated switches designed to close the appro-
priate detonator circuits upon rapid decelera-
tion of the projectile. This device is particu-
larly useful when a large number of duds is due
to the failure of electric components other than
the power supply. A simple form of this device
(for the T-6 fuze) is illustrated in Figure 11,
where two leaf springs equipped with silver
contacts are mounted on the forward plate of
the switch mechanism and are arranged so as
to provide direct connection between the deto-

Figure 9. Diagrammatic illustration of self-
destruction mechanism.

detonation, however, did not become a part of
the formal military requirements (Ordnance
Committee Minutes) until development of mor-
tar shell fuzes was initiated (see Section 1.1).
The arguments for the incorporation of this
impact element is that in case of a failure of
some electric component the fuze would still
detonate the charge upon collision with the
target or with the ground. This would serve
both to increase the effectiveness of the weapon
and to increase security by decreasing the
number of duds. The objections to the use of
an impact detonator are the greater complexity
required in the fuze and the very great danger
present when mechanical detonators are used,
particularly so since the dud clearance prob-
lems encountered by our services were very
severe.

Figure 10. Self-destruction element for T-5
fuze.

nator and the battery.7 Only 100 of these
switches were built before the work on the
battery fuzes was terminated. The tests per-
formed by the Army showed excellent results,
with the switches operating properly upon
ground impact down to an impact angle of 15
degrees. Such mechanisms can be made as sen-
sitive as desired, and the only limitation is the
drag of the projectile in flight. In the case of
the M-8 rocket, this drag occasionally reached
maximum values of 18#.

4 3 MECHANICAL DESIGN OF

PROXIMITY FUZES

4 31 Battery Fuzes

The first of the fuzes developed under the
auspices of this division, for which definite

SECRET

176

MECHANICAL DESIGN

mechanical characteristics were specified by
the services, were the battery fuzes for the M-8
rocket. The photoelectric and the radio fuzes
were to be made interchangeable in all external
respects. They were to fit a 3-in. fuze well 5 in.
deep, and the nose contour was to be a continu-

Figure 11. Impact detonator for T-6 fuze.

arming switches are contained in a metal hous-
ing. The booster charge, which is a block of
tetryl y2 in. long and roughly 3 in. in diameter,
is contained in an appropriate compartment in
the fuze housing directly below the arming
switch.

The main features of the arming mechanism
for this series of fuzes are as follows. The
arming mechanism proper is energized by the
acceleration of the rocket. This acceleration
acts upon a small lead weight fastened to an
escapement wheel which is retained by a 75-g
coil spring. The motion of this wheel is con-
trolled by a flutter bar. The arrangement can
be seen in Figure 5. If the mechanism is acted
on by an acceleration greater than 75 g for a
time longer than 0.15 sec, the lead weight
reaches a position roughly 90 degrees from its
starting point. This permits the operation of
a spring-driven switch that closes the A and B
power supply contacts, also shown in Figure 5.
Upon cessation of the acceleration, the 75-g
spring acting on the escapement wheel reverses
its motion and moves a powder train inter-
rupter carrying a tetryl lead into the armed

ation of the rocket ogive. A photograph of the
T-5 fuze is shown as Figure 12. The T-4 photo-
electric fuze, which is described in Volume 3
(Division 4 STR) used the same battery,
switch, and housing as the T-5.

Since the battery was to be easily replace-
able, each of the fuzes was broken down into
three separate components: the head or elec-
tronic assembly, the battery, and the safety and
arming mechanism. The three parts were ar-
ranged to be connected by plugs and sockets.

The mechanical design of the head is rather
simple. In the case of the radio fuze, the nose
is made of mica-filled phenolic into which a
small antenna cap is molded. In some of the
fuzes this cap was spattered onto the surface
of the phenolic. The nose is hollow and contains
an oscillator block supported by a metal shelf
below which are mounted the amplifier com-
ponents. The base of this head consists of an
insulating plate provided with pins which fit
into corresponding socket holes in the top plate
of the battery. The arming mechanism and

Figure 12. T-5 fuze, components and assembly.

From left to right: electronic assembly desig-
nated as MC-382; battery power supply desig-
nated as BA-75; arming switch designated as
SW-200; housing which contains tetryl booster;
partial assembly of fuze; fuze completely assem-
bled ready for installation in rocket.

position. Mounted on this interrupter bar is a
small switch element that closes the detonator
circuit near the end of the bar’s travel. In this
manner the fuze is not fully armed either elec-
trically or mechanically until long after the
cessation of acceleration. In most of the switches

MECHANICAL DESIGN OF PROXIMITY FUZES

177

built, this amounted to roughly 0.8 sec after the
launching of the rocket.

A note about switch contacts should be made
at this point. In many of the switches which
preceded the production model described above,
the contacts were mounted on leaf springs, as
is the common practice in many relays and tele-
phone jacks. It was soon discovered that these
“pressure” contacts were extremely micro-
phonic and caused malfunction of the fuzes
because of the resultant electric noise. The
rotary-type of radio switch that was finally
adopted was far better in this respect. This
was probably due to the fact that the contact
pressures were somewhat greater and that the
springs were extremely short, resulting in ex-
tremely rigid contact assemblies. Other types
of contacts, particularly of the wedge-type,
were also tested, but the availability of the type
shown resulted in their exclusive use. Work on
the engineering and the production phases of
these arming mechanisms was done by the
Globe Union Company of Milwaukee. Many
variations of these switches were built by this
company. They differed in the arming time, the
value of acceleration for operation, the pres-
ence of impact detonating switch elements, the
incorporation of SD, and many variations in
the electric circuit.

The detonators for the fuzes were inserted
into the arming switches through an opening
in the top plate. This operation could be per-
formed after the switch was completely assem-
bled and tested. At the request of Army Ord-
nance an additional safety feature was later
added which consisted of a key passing through
the top plate of the switch. (See Figure 5.)
The key was so arranged that if the mecha-
nism, for some reason, began its arming cycle,
the key could not be removed. This in turn pre-
vented the switch from being plugged into the
battery. If the arming mechanism was in the
correct and safe position, the key could be
easily removed and discarded. Of the hundreds
of thousands of switches made, tested, and
used, no case of malfunction resulting in an
accident was reported.

Since the assembled fuze had to be capable
of withstanding an acceleration of several hun-
dred g, the components were required to pass
a centrifuge test at l,000p. Large centrifuges

were built for these tests by the National
Bureau of Standards [NBS] and by the vari-
ous contractors involved (see Figure 13).

Generator Fuzes for Rockets
and Bombs

Early RRLG Fuze for Rocket Application

As mentioned elsewhere in this report the
shortcomings of the battery were quite obvious
to all concerned, and, shortly after the battery
fuzes went into production, work was started
on the use of air-driven generators for power
supply. Since the efforts at that time were di-

FlGURE 13. Centrifuge for testing T-5 fuze.

rected toward the development of a fuze for
the M-8 rocket, the generator and its driving
system was designed to fit into the general pat-
tern of the T-5 fuze; there was a strong effort
made to use as many components of that fuze
in its successor as possible. The obvious solution
was to mount the generator below the head,
in the space originally occupied by the battery,
and to drive it by means of an insulating shaft
connected to a windmill in the nose. Such an
arrangement is shown in the radio rocket lon-
gitudinal generator [RRLG] fuze shown in
Figure 14. The antenna still consisted of a
small metal cap directly under the vane, all of
the vanes used in these fuzes were of Bakelite.

SECRET

178

MECHANICAL DESIGN

The shaft connecting the windmill to the gen-
erator had to be of an insulating material so
as not to short-circuit the antenna system.
Cloth-filled Bakelite was found most suitable
for this purpose.

At this time precision ball-bearings were not
available in quantity, and in the first attempts
to get satisfactory bearings for the high speeds
involved, porous bronze (Oilite) bearings with
steel shafts were used. These were acceptable
in the generator but did not operate properly

In connection with the RRLG the question
arose as to whether the old plug-in arming
system should be retained, operated only by set-
back, or whether advantage should be taken of
the vane action and the arming be made depend-
ent on both acceleration and air travel. Since
the emphasis on the safety of the proximity
fuzes at the beginning of World War II was per-
haps inordinately large, it was decided to aban-
don the pure setback mechanism and to employ
a safety device which would not arm unless

FILTER CONDENSER

GENERATOR COVER

ENCASING CAN

SQUIB ROTOR

RF-AF SUBASSEMBLY
POTTED

ASSEMBLED POWER SUPPLY
AND ARMING MECHANISM

OSCILLATOR BLOCK AND
SHIELD PLATE

COMPLETE ASSEMBLY

GENERATOR ROTOR
SET BACK LOCK

SAFETY PLATE TETRYL CUP

AMPLIFIER ASSEMBLY

RF-AF SUB ASSEMBLIES

COMPLETE RF-AF HEAD

T-12 (RRLG) ROCKET RADIO FUZE
GENERATOR POWERED

GENERATOR SET BACK RELEASE
ARMING GEAR TRAIN RECTIFIER
SELF DESTRUCTION

Figure 14. T-12 fuze, complete assembly and principal components.

as nose bearings, primarily because of the
thrust on the windmill and the large amounts
of unbalance present. Home-made ball-bearing
races were utilized with success (Figure 15).

In the rocket application, for which the first
generator fuze was designed, the air velocities
were limited to a rather narrow range between
800 and 1,500 fps, and no difficulties were ex-
perienced with excessive speeds of the gen-
erators. This was not the case later in bomb
fuzes.

acted upon by both air impact and acceleration.
The RRLG fuze, or T-12 fuze (Ordnance De-
partment nomenclature) , was not manufactured
on a large scale because, at approximately the
time the design was completed, further work
on the M-8 rocket was stopped. The general
philosophy, however, of combining air drive
with setback was carried over into all the later
fuzes. Biblographical references dealing with
this fuze are NBS reports on RRLG or T-12 and
reports of the Rudolph Wurlitzer Company.74

| SECRET

MECHANICAL DESIGN OF PROXIMITY FUZES

179

T-50 Bomb Fuze

Overall Design Details. The next requirement
for a proximity fuze was for a generator-
operated fuze for bomb use. Since the head and
the generator as used in the RRLG appeared
to be satisfactory, they were incorporated into

facturers of Alnico rotors and of the fuzes (see
Figure 17). The rotating system of this ultra
centrifuge is an air-supported turbine rotor
capable of speeds in excess of 120,000 rpm.

The original windmills had three blades of
2-in. overall diameter ; however, when they
were released at high altitudes and at low plane
velocities, the rotational speeds did not repeat
well. Therefore, the design was changed to a
2V2-im windmill that had considerably larger
power output, resulting in more reproducible
speed. Since the bomb velocity varied from ap-
proximately 200 to 900 fps, the speed of the
vane varied over a corresponding range. This
resulted in extremely high top speeds; since

GENERATOR PROPELLER

Figure 15. Vane and vane bearings for gen-
erator-powered bomb fuze.

the T-50 series of fuzes but with some changes
in the arming system and the overall exterior
shape in order to adopt them to bomb use. The
nose fuze well of most of the American bombs
used a 2-in. thread ; therefore, an adapter case,
which because of its general physical appear-
ance became known as the “potato masher,”
was designed to house the entire mechanism.

A photograph of a cutaway of a typical T-50
fuze is shown in Figure 16. As can be seen,
the general arrangement is similar to that of
the RRLG fuze in that the electronic compo-
nents of the nose are followed by the power
supply and the arming system. The rectifiers
for the power supply are mounted in the Bake-
lite housing surrounding the reduction gear,
while the filter and firing condensers are made
in a tubular shape and mounted in the space
surrounding the low-speed arming shaft. The
vanes, or windmills, originally used in the T-50
were made of either cotton flock or rag-filled
phenolic materials. It was found that such
vanes could withstand rotational speeds up to
80,000 rpm without bursting. Some of the cast
Alnico generator rotors could not withstand
such high speeds, and it was necessary to per-
form a large amount of high-speed testing. An
ultra centrifuge was developed for this purpose
at NBS and was adopted by many of the manu-

TRANSMITTER

RECEIVER

POWOER TRAIN

GENERATOR

FIRING UNIT

Figure 16.
fuze.

Cutaway of typical T-50 type bomb

dynamic balancing was not employed in the
initial production of these fuzes, great difficul-
ties arose due to the failures of bearings and
the presence of microphonic noise.

Several lines of attack were followed to over-
come these difficulties. One concerned the use
of interchangeable windmills that could be eas-

CRET

180

MECHANICAL DESIGN

ily changed in the field in order to select the
type most suited to the plane speed and the
bombing altitude. Another involved the bal-
ancing of the vanes to eliminate vibration and
thus permit the use of a single high-speed wind-
mill for all applications. It was found that the
interchangeable units could not be easily bal-
anced, and this method was soon abandoned.

Figure 17. Ultra centrifuge for testing rotors
for T-50 type fuzes.

One interesting by-product of the plan to use
interchangeable turbines was the special T-50
shipping can with a special container built into
its cover for one or two spare units. This addi-
tional space later proved itself very convenient
for packaging the T-2 extended arming device
which, in fact, was specifically designed to fit
this package.

Dynamic balancing will be discussed at the
end of this chapter (see Section 4.6).

In the first bomb fuzes the antenna still con-
sisted of a small streamlined cap directly below
the vanes, but for electrical reasons it was soon
changed to a thin ring approximately 3 in. in
diameter that was supported by four buttresses
extending from the nose section, as seen in
Figure 18. A short time later the ring was
lengthened to approximately % in. (see Fig-
ure 16) and this was the antenna that, with
minor variations, was carried into all the later
fuzes of this general type. (See Section 2.7.6
for electrical reasons for increasing length of
ring.) This longer antenna ring also performed
several useful mechanical services. It acted as
a guard for the vanes and at the same time pro-
vided a convenient anchorage for the vane
locking pins and extended arming devices.

The enclosing of the windmill in a long an-
tenna ring also gave rise to the possibility of
using metal vanes. Successful tests were made,

and shortly thereafter most of the manufac-
turers changed to the use of stamped steel
10-bladed windmills. Some difficulty was ex-
perienced with metal fatigue and breakage of
the blades, but this was rectified by the use
of ribbing at the thin section near the root of
each blade (see Figure 19). These windmills
were also dynamically balanced in production.
Until very near the end of the T-50 program
the vane shafts were equipped with ball bear-
ings of the type shown in Figures 15 and 18B.
The races were machined in steel and case-
hardened. The balls were of the quality used in
precision bearings. Because of the absence of
thrust and of the inherently better balance, the
generator bearings were of the simple sleeve
variety. The shafts were of stainless steel, while
the bearings were of porous bronze, commer-
cially known as Oilite. Near the end of World

Figure 18A. Photograph of assembled T-50
type bomb fuze.

War II most of the manufacturers began to
use precision ball bearings, both for the tur-
bines and in the generators.

In the original design of the T-50, the cou-
pling shaft, that is, the shaft coupling the
windmill to the generator, was loosely coupled
at both ends. The engineers of the General
Electric Company suggested and experimented
with a design in which the coupling shaft was
rigidly attached to the windmill but was loosely
coupled only at its bottom end. This required
only one ball bearing at the nose instead of two.
The design was generally adopted and its use
resulted in a simpler, more rigid, and more
economical assembly.17 It is shown in Figure 16.

Experiments at Bowen75 and at the National
Bureau of Standards18 showed that noise was

MECHANICAL DESIGN OF PROXIMITY FUZES

181

still caused by the looseness of the coupling
between the insulating shaft and the generator.
Experiments were performed on the use of
rubber and other flexible materials as vibration
absorbers at this point. A final design was
evolved in which the generator was driven
through a tight-fitting rubber coupling. This
also served to minimize the rotational oscilla-
tions of the generator rotor that had caused
phase modulation in the generator output. This,
in turn, modulated the voltage output of the

wires were employed to hold the vanes and the
arming mechanisms in the “safe” condition
(see Figure 20). After release, the windmills
were required to make a definite number of
turns to arm the fuzes. The electric arming of
the generator fuzes was considerably simpler
than that of the battery fuze, since no A or B
switches were required. In the original RRLG
rocket fuze, SD was a requisite, and the gear
train was, therefore, arranged to continue its
operation after the explosive train was aligned.

Oscillator block

Amplifier

Generator

Rectifier

Filter condenser

Contacts to detonator

Detonator

Tetryl plate

Windmill

Vane bearing assembly
Drive shaft

N Antenna (in later models, antenna extended forward to
enclose vane)

P Insulating support for antenna
Q Fuze housing (“potato masher”)

R Speed reducing gears
S Lugs for wrench

T Low-speed drive shaft for arming mechanism
U Locking pin for detonator rotor

W Detonator rotor (arming consists of rotation of this piece
into proper position)

X Booster cup

Figure 18B. Sectionalized drawing of T-50 type bomb fuze. Same general arrangement of parts used
for all ring-type bomb fuzes.

power supply when the latter was operated on
the steep part of its voltage-versus-speed regu-
lation curve.

Some further reduction in mechanical noise
could have been obtained by dynamic balancing
of the generator rotors, but because this would
have necessitated major changes in the assem-
bly, it was not resorted to in the T-50 and T-30
series of fuzes.

The Arming System. Since bombs experience
no acceleration of large magnitude, arming

An SD contact was arranged to close the det-
onator circuit at the desired time after arming.
This feature of a continually running gear
train was carried over into the bomb fuze,
since the possibility of converting them back
into rocket fuzes was always present.

From the point of view of noise and micro-
phonics, it would have been better to design
the gear train so that it would be disconnected
from the vane at arming, and several methods
of doing this were, in fact, suggested, but the

SECRET

182

MECHANICAL DESIGN

possibility of requiring SD and the availability
of the gear train (from the RRLG program)
together with the ever present pressure of
time kept the mechanism as it was. The first
gear trains used in the T-50 series of bombs

Figure 19. Unfinished windmill for T-50 bomb
fuze, showing flutings for increased rigidity.

were of the planetary-differential type. A pho-
tograph of this gear train is shown in Figure
21. This differential gear appeared particularly
desirable for these mass-produced fuzes be-
cause of its simplicity and cheapness. It was
designed originally for the RRLG rocket fuze,
for which speeds in excess of 80,000 rpm were
not expected, and no serious trouble with noise
or short life was anticipated. This, however, did
not prove to be the case in the bomb applica-
tion. These gears suffered from several grave
defects. One was their short life under high
speeds of operation, and the other was the
large amount of noise and vibration they intro-
duced into the fuze.3 In order to overcome these
difficulties, a worm type of gear reduction,
which fitted the space allotted to the differential
gear, was designed (see Figure 22). The only
change necessary in the fuze for its adoption
was a change in the generator shaft, which
now incorporated a worm at the first step of
the gear reduction. These worm gears, which
were engineered by the Globe Union Company51
and produced by them and several other con-
tactors, were used in all of the bomb fuzes
with excellent results. The overall reduction in
speed from the windmill to the arming shaft
was approximately 5,800 to 1.

A rather radical departure from the previous

arming systems was introduced into the method
of interrupting the powder train. Instead of
keeping the electric detonator in a fixed posi-
tion and moving a slider containing one of the
other powder train elements, it was decided
that considerable increase in simplicity and de-
pendability could be achieved by moving the
electric detonator itself so as to interrupt both
the explosive train and the electric firing cir-
cuit. A small Bakelite drum was arranged to
be driven by the low-speed shaft of the gear
train. This drum carried the detonator with

Figure 20. Use of arming wire and locking pin
to prevent vane rotation.

its two electric contacts and a small transfer
pin which acted as a coupling between the drive
shaft and the drum and also served as a lock
for the detonator assembly when in the armed
position. Since the gear train ran continually
before and after arming, SD could be achieved
by the insertion of a special washer under the
detonator drum in order to ground the thyra-

SECRET

MECHANICAL DESIGN OF PROXIMITY FUZES

183

tron plate circuit and thus fire the detona-
tor.

Because of the direct relation between the
vane speed and the velocity of air travel, the

Figure 21. Planetary-type speed-reducing de-
vice for T-50 fuzes.

the tetryl lead in the safe or unarmed position.
This limited the unarmed angle setting, called
the arming angle, to the range of more than
60 degrees and less than 300 degrees. Angles
from 100 to 180 degrees were most commonly
used in practice.

The electric connections to the detonator were
made through two phosphor bronze or beryl-
lium-copper leaf springs mounted in the det-
onator rotor housing. These silver-plated
springs made contact to two small silver-plated
screw heads, under which the detonator leads
were fastened. The system of contacts described
above had several serious faults: the springs
could be easily deformed in handling so as to
result in poor contact, and the transfer pin
had to be rather carefully made because if it
failed to snap out of the slot in the shaft and
lock the detonator rotor into its armed position,
the rotor would turn through the armed posi-
tion and cause the fuze to be a dud.

The exact length of air travel to arming
could not be exactly preset because the manu-
facturing tolerances in this assembly were such
that small erorrs could cause large differences

PROPELLER GENERATOR SLOW SPEED SHAFT SAFETY PLATE

ROTOR

t - ' Lw,. V <7* ;

COUPLING SHAFT GEAR TRAIN DETONATOR ROTOR TETRYL CUP

Figure 22. Arming mechanism of T-50 type fuzes with worm gear.

distance through which a bomb fell before arm-
ing was easily controllable by a change in the
angular setting of the detonator rotor. The only
limitation on this was the fact that the det-
onator had to be at least 60 degrees away from

in the arming distance. Some of the fuzes pro-
duced near the end of the program were set by
manual or automatic counting of windmill
turns. This difficulty was not anticipated in the
design because the need for precision arming

184

MECHANICAL DESIGN

of the proximity fuze was not expected. It was
thought that merely delaying the arming for
an approximate distance would be sufficient.
The using services, however, laid down rather
stringent requirements during, the course of
the development program, for the minimum
and maximum limits of safe air travel, and
several minor modifications in the arming sys-
tems were introduced as a result.

A removable safety pin similar to the key
of the T-5 switch was added to the arming sys-
tem of the T-50. This pin had to be manually

Figure 23. Safety pin installed in T-50 type
fuzes. Pin is removed before fuze is inserted in
fuze well.

later manufactured as the T-51 fuze by the
Zenith Radio Corporation (see Figures 26 and
27). 71 The final model of T-51, which carries a
bracket for the vane locking pin, is shown in
Figure 5 of Chapter 1.

Several types of dipoles were experimented
with at the National Bureau of Standards. One
was the metal type molded into the antenna
head, and the other consisted of plastic dipoles,
made integral with the head, over which a con-
ducting surface of metal was either plated or
spattered. Metal dipoles were used by the
Zenith Corporation in their production.

T-82 Bomb Fuze

A markedly different bomb fuze, the T-82,
was developed by the Westinghouse Company.68

Figure 24. Details of arming safety pin de-
vice.

removed before the fuze could be screwed into
the bomb well. This pin indicated that the det-
onator rotor was in the safe position, and it
could also be used as a later check if, for some
reason, the fuze had to be removed from the
bomb after a flight. The details of this safety
system are shown in Figures 23 and 24.

T-51 Bomb Fuze

A modification of the T-50 fuze was made by
changing the antenna to one of the dipole
variety (see Figure 25). The mechanical ar-
rangement was almost exactly identical to that
of the T-50 except that plastic windmills were
used to the end of production. This fuze was

(See Figure 28.) In an effort to overcome the
vibration difficulties encountered in the design
of the T-50, the rotating system of this fuze
was located in its base so that it could be nearly
totally enclosed by the fuze well. By using a
radial flow turbine mounted directly on the
shaft of the generator and supporting the
whole high-speed assembly by a metal casting
mounted in the nose of the bomb, an extremely
rigid and quiet mechanical design was achieved.
The air to drive the turbine was conducted
through the electronic components by a central
duct and exhausted through two wide ports
near the base of the fuze. The main body of the
fuze was made of mica-filled phenolic and was

MECHANICAL DESIGN OF PROXIMITY FUZES

185

Figure 25. Early model of T-51 fuze.

gether by four screws. The gear reduction was
of a worm type similar to that used in the T-50,
and the arming system was identical.

No dynamic balancing was employed in this

Figure 27. Assembly and principal components
of T-51 fuze.

connected electrically to the power supply and
the arming system by means of a multiple pin
plug. These two main assemblies were held to-

fuze, but precision ball bearings were used in
all models. In order to limit the top speed of
the turbine, automatic speed regulation of sev-
eral kinds was tried, and it was found that, by

ASSEMBLY ASSEMBLY HOUSING

GENERATOR

ASSEMBLY

£ #

ADAPTER ROTOR TETRYL PLATE

rrrrrT“;-"'-r:!i

Figure 28. Assembly and principal components
of T-82 fuze.

making the blades of the turbine of spring
steel, they could be made to change their curva-
ture, or pitch, when acted upon by both cen-
trifugal force and air pressure. The production

Figure 26. Later model of T-51 fuze.

186

MECHANICAL DESIGN

model of the fuze employed four rigid and four
flexible plates, shown in Figure 28. It was
found in practice, however, that the range of
speeds over which this fuze operated did not
result in any speed regulation of this particular
turbine, but the top speeds did not cause any
trouble, because of the excellent bearings
used.

train was, therefore, developed at the National
Bureau of Standards, which was again de-
signed for production and produced by the
Globe Union Company. Figure 29 shows the
construction of this gear train, and again only
the main features of its operation will be
stated. For proper arming the mechanism re-
quires a sustained acceleration of more than

Figure 29. Arming device for T-30 and T-2004 fuzes. A, parts in their normal position; B, inertia
element in position assumed during setback.

Generator Rocket Fuze
T-30 (and T-2004)

As mentioned previously, work on the M-8
rocket was discontinued, and emphasis was
placed on the use of the fin-stabilized Navy
rockets developed at the California Institute of
Technology. It was quite apparent that with
slight modifications the T-50 bomb fuze would
serve excellently on these projectiles. With a
minor change in the vane pitch, the fuze could
have been used “as is,” but the presence of
reasonably large values of acceleration offered
attractive possibilities of increasing the safety
of the arming mechanism. A new type of gear

lOp occurring simultaneously with the rapid
motion of the fuze through air for 300 ft. The
windmill is locked by the usual arming wire,
when the rocket is in its launcher. If this arm-
ing wire were prematurely withdrawn, the
windmill would start rotating, but due to the
absence of acceleration the mechanism would
jam, one of the brass gears would strip, and the
fuze would become a dud if fired.

The arming system was further designed so
that the mechanical and electric arming was
not completed until after the cessation of accel-
eration. This was done to prevent the fuzes
from being set off at some point beyond the
300 ft by the burning of the propellant. The

MECHANICAL DESIGN OF PROXIMITY FUZES

187

RC arming delay which followed the comple-
tion of the mechanical arming cycle increased
this safety still further.

Since this gear train was so designed that the

Figure 30. Doughnut arming device installed
in fuze.

low-speed shaft did not rotate after the comple-
tion of mechanical arming, no SD was incor-
porated into the T-30 and the T-2004 fuzes. For
the same reason, the detonator rotors were not
provided with a transfer pin but were perma-
nently locked to the low-speed shaft.

The T-30 and the T-2004 fuzes were pri-
marily designed for stop-gap use, while the de-
velopment of special rocket fuzes was in prog-
ress (see Section 4.3.4 on T-2005) and the
setback gear train was designed with this in
mind. The major objection to the modified bomb
fuzes as rocket fuzes was their size, which meas-
urably increased the drag on the missile. The
use of arming wires in rocket fuzes was consid-

ered objectionable by NBS engineers, and the
employment of a setback mechanism that could
not be inspected from the outside and that re-
sulted in a dud in case of malfunction of the
arming wire was not considered an elegant
solution.

The British Air Forces requested Division 4
to design an arming mechanism that would
convert a T-50 into a rocket fuze but that would
keep the vane from turning and release it only
by the action of setback. No arming wires were
to be required and no loose components, such as

Figure 31. Doughnut arming device for rocket
fuzes. A, plunger in armed position; B, plunger
in unarmed position where it prevents rotation
of vanes.

pins, were to be released in flight. Accordingly,
a “doughnut” arming mechanism was devel-
oped.41 This mechanism, shown in Figures 30
and 31, fitted inside the antenna and held the
vanes in the locked position by a small pin. A

SECRET

188

MECHANICAL DESIGN

flutter type of mechanism was enclosed in the
ring and, when subjected to an acceleration of
more than 10 # for more than 14 sec, caused the
release of the vanes. The engineering and pro-
duction of this unit was done by the Transition
Office of NDRC and by the Solar Aircraft Com-
pany of California. An interesting feature of
this device is the circular flutter weight, with
its center of gravity on the center line of the
fuze. This made the mechanism operable in the
presence of slow rocket spin (about 1,000 rpm,
the spin rate of one of the British rockets for
which the mechanism was designed). The use
of the ring-shaped weight enabled its designers
to obtain a large moment of inertia with a
minimum of total mass. This device was able to
pass the standard jolt test without difficulty.

Figure 32. Front view of high-# centrifuge for
testing mortar fuzes.

4 3 4 Miniature Fuzes for Trench Mortars
and Rockets

Fuzes, T-132 and T-171

When Division 4 undertook the development
of a generator proximity fuze for trench-mor-

tar use, completely new problems of mechanical
design arose. The accelerations experienced by
a mortar shell are of entirely different magni-
tude from those to which the physicists and
engineers were accustomed in their work on

Figure 33. Side view of high-# centrifuge for
testing mortar fuzes.

bombs and rockets. An 81-mm mortar shell is
fired from its gun with varying accelerations up
to 6,000#. The resulting stresses obviously re-
quired a new approach to the mechanical design
of the proximity fuze. This point in the fuze
program represented a welcomed opportunity
for incorporating into the new fuze a great
many of the suggestions and ideas which were
gathered in the previous work.

New testing techniques had to be developed
for this work. Although it is practically impos-
sible to duplicate the gun accelerations in the
laboratory, a close approximation can be made
by using a special centrifuge. Accordingly,
NBS designed and built a high-# centrifuge
capable of testing complete mortar fuzes at
accelerations up to 15,000#. Photographs of
this equipment are shown in Figures 32 and

MECHANICAL DESIGN OF PROXIMITY FUZES

189

33. Two Dural arms, such as are used in the
machine, can be seen in the photographs.

The requirement for extreme compactness
presented, besides the mechanical problems, the
problem of securing a sufficiently large antenna
to insure adequate r-f loading. In order to ac-
complish this, the usual arrangement of the
antenna and ground of the proximity fuze was
reversed. It was decided to make the body of
the fuze the ground, and to use the vehicle as
the antenna. This, of course, is merely a jug-

Figure 34. Diagram showing arrangement of
principal components in T-132 and T-171 fuzes.

gling of words, but it helps to explain how, for
a fuze of a given size, a better arrangement can
be made by locating the antenna insulator di-
rectly ahead of the nose of the projectile and
mounting as many of the mechanical and elec-
tronic components of the fuze as possible ahead
of this antenna “break.” A diagram of the
general arrangement of the resulting fuze is
shown in Figure 34.

The possibility of using a base generator, as
was done in the case of the T-82, was seriously
considered but was discarded because of the
considerable space occupied by the air passages.
Accordingly, the turbine and the generator
were mounted in the forward end of the fuze.
The decision to do this was further strength-
ened by the requirements that the fuze operate
at very low airspeeds. In the case of a 0 charge,
the 81-mm shell leaves the gun at approxi-
mately 150 fps, making the successful opera-
tion of the turbine difficult. More will be said
about this matter in connection with the T-172.

The generators used in the experimental
trench mortar fuzes were originally identical
with those designed by Zenith for the T-50,
with the exception that the six corners of the
stator were machined off, giving a circular
stator with an outside diameter of 2 in. It was
found that the removal of the corners did not
reduce the output of the generators. This six-
coil design was adopted by the Globe Union
Company in their production of the T-132.50
The Wurlitzer Company, however, because of
their experience with the wave-wound T-50
generators decided to experiment with a double
snake-wound generator of somewhat smaller
size (see Figure 35). This design, when used
with a voltage doubling rectifier circuit, proved
quite satisfactory.

The problem of supporting the coils and

A B

Figure 35. Generator stator for T-171 fuze
(right) shown in comparison with stator for
T-50 fuze (left).

laminations against the force of setback was
met by centrifugal potting of the stator, using
a special high-temperature potting compound
(see Section 4.7). The stator of the generator
was enclosed in a thin metal shell, and the
whole assembly was rotated about the central
axis of the generator at approximately 7,000
rpm. A measured quantity of the hot potting
mixture was poured into the generator frame.

SECRET \

190

MECHANICAL DESIGN

The centrifugal action forced the liquid to
spread into the space around the coils and form
a cylindrical inner surface just back of the
pole faces. The material was cooled while still
spinning at the high rate.

The dynamic balancing of the high-speed
rotating system eased the problem of the bear-
ing design very greatly. Precision bearings

Figure 36. Protective cover and arming pin for
T-132 fuze.

were incorporated into only a small number of
fuzes. It was found that the New Departure
R-3 (V2 in. OD, %6 ID) was capable of with-
standing a static thrust of nearly 1,000 lb and
then was able to operate at 100,000 rpm for
several minutes without failure. The procure-
ment problem was still very serious and since
experiments indicated that sleeve bearings,
particularly of the Oilite-type, could be em-
ployed satisfactorily, the Globe Union Com-
pany did considerable research on their use.
The engineers of the Allis-Chalmers Company
in Milwaukee urged the adoption of rubber
mounting of the bearings, since their experi-
ence with the high-speed rotating machinery

indicated that some form of damping was re-
quired for these bearings. The Globe Union
Company adopted this suggestion, and the use
of rubber mounting for the sleeve bearings was
standard in all of their production of the T-132.

In the summer of 1945 the University of
California was asked by the Transitions Office
of NDRC to pursue further the research on
the bearings and rotating components for the
mortar fuzes. This group found that shafts
with extremely hard surfaces were suitable for
this service.60 The National Bureau of Stand-
ards also conducted research in the same field
and had excellent results with bearing assem-
blies in which both the shaft and bearing were
made of identical and very hard materials.
Some Nitralloy bearings mounted in rubber
were run at speeds in excess of 75,000 rpm con-
tinuously for an hour without failure. This was
all the more amazing, since the bearings were
not lubricated in any manner whatever.25

The high-speed joints of the T-50 were elimi-
nated by the single unit rotating assembly con-
sisting of a turbine, the generator rotor, and
the high-speed shaft. Since dynamic balancing
required the removal of metal in two planes,
a brass disk of approximately %6- in. thickness
was fastened below the Alnico rotor. The under
side of the turbine and this brass disk provided
two convenient surfaces for the easy removal of
mass. A special dynamic balancing machine for
this purpose will be described in Section
4.6.

The arming system of the T-132 fuze in the
original form was to be operated by the impact
of air so that the fuze would arm after the
maximum possible air travel. Calculations
showed this to be approximately 400 yd. This
would permit the fuze to be fired at 0 incre-
ments and 45-degree elevation, with the arming
occurring a very short distance before im-
pact.

A manually removable safety pin was pro-
vided, as shown in Figure 34. This pin was in-
tended to prevent accidental arming of the fuze
but had to be removed and thrown away in
accordance with the customary use of the
standard trench-mortar mechanical fuzes. In
the final T-132 designs, a protective plastic
cover was shrunk over the front end of the

MECHANICAL DESIGN OF PROXIMITY FUZES

191

fuze, and the removal of the pin and the cover
was accomplished by one motion of the hand.
Photographs of the arrangement are shown in
Figure 36.

Several departures from the bomb and rocket
arming techniques were made in the details of
the arming system. The detonator was still
carried in a detonator rotor mounted above an
interrupter plate. It was maintained in its safe

Figure 37. Jolt machine for testing proximity
fuzes.

position by an arming shaft, which was with-
drawn by the action of a screw driven by the
turbine through a reduction gear train. Its
movement, however, was not slow as in the case
of the bomb and rocket fuzes. Instead the deto-
nator rotor was snapped into position by a coil
spring. In the T-132 the arming shaft was made
of Bakelite so as not to short-circuit the an-
tenna system. It was enclosed in a metal shield-
ing tube for as great a part of its length as pos-
sible. In the T-171 a metal shaft was used; a
short circuit at the antenna was prevented by
a snap-out motion of this shaft, rapidly with-
drawing it from the antenna insulator at the
moment of arming.

The original design of the detonator rotor
provided an additional space for a mechanical
impact detonator element, but because of the
very great danger in using this element, it was
not built into the first production of the fuzes.
Instead the space was occupied by a double-
element safety pin which held the rotor in a
safe position unless released by a sustained
acceleration of over 1,000# (see Figures 6 and
38). This additional safety was necessitated by
the requirement that the mortar fuzes be able
to pass the jolt test. This test consists of mount-
ing the fuzes by their base threads into a
machine-driven arm that is subjected to a free
fall of approximately 3 in. onto a rather hard
surface for a total of 5,250 drops (see Figure
37). The fuzes were held in each of three posi-
tions for 1,750 drops each. It was found that
the first T-132 and T-171 fuzes built would not
pass this test but failed by breaking at the an-
tenna insulator. With the arming system as
originally designed, this resulted in the with-
drawal of the arming shaft and the rotation of
the detonator into the armed position. Since
the electric detonators are quite safe against
mechanical shock at handling, this did not nec-
essarily represent a serious hazard. However,
the addition of the double-element safety pin
eliminated even this danger by keeping the
rotor in the safe position in case of accidental
withdrawal of the arming shaft.

Still another version of the arming mecha-
nism for the T-132 and T-171 series of fuzes was
designed just before the end of World War II.
In an effort to increase the unarmed air travel,
a clock mechanism was substituted for the gear
train. This clock mechanism, shown in Figure
38, was mechanically coupled to the detonator,
and the whole assembly occupied the space
originally filled by the detonator rotor. The
detonator was held in the safe position by the
double-element setback pin described above for
the original rotor. Upon its release the detona-
tor was moved into position and was electrically
and mechanically armed 10 sec after firing.
This resulted in an automatic increase in the
safe air travel when the shells were fired with
the greater number of increments, because the
air travel to arming now was directly propor-
tional to the velocity of the shell.

EGRET

192

MECHANICAL DESIGN

Another great advantage of using this clock
rotor was the elimination of the gear train and
arming shaft. This permitted the power supply
to be completely isolated mechanically from the
rest of the fuze so that excellent sealing of the
electronic components against moisture became
possible. The elimination of the gear train with

Figure 38. Clock mechanism for arming mortar
fuzes. Pin assembly in lower right-hand corner
is common to all fuzes developed for mortar
shells. It is a double-element inertia device.

its arming shaft also meant considerable re-
duction in mechanical vibration and noise.

Raymond Engineering Laboratories were
asked to do the engineering and the experi-
mental production of the clock rotors.57 Several
completely satisfactory working samples were
received from them shortly before the conclu-
sion of hostilities.

Another expedient which was experimented
with for delaying the arming of the fuze for
as long a time as possible was the use of a
small dashpot switch ; a cross section of this is
shown in Figure 39. The switch would be placed
into the detonator circuit and would normally
be kept closed by the coil spring. The detonator
would, however, be kept unarmed by the regu-
lar clock rotor. Upon firing, the sliding piston
contact would move back for a distance propor-
tional to the time integral of the acceleration,
and, upon cessation of acceleration, it would be

moved forward by the spring, until it finally
acted to close the detonator circuit. Since the
time of the return stroke was dependent on the
length of the piston travel, automatically vari-
able arming time could be secured that would
be controlled by the amount of explosive charge
used in firing the shell. A considerable amount
of experimental work was done on this device,
but it was found that the difficulties in its con-
struction made it impractical. It was expected,
for instance, that the effects of temperature
upon the viscosity of the fluid, one of the sili-
cones, would be largely canceled out because of
the double action of the piston; that is, that
temperature effect on the down stroke and on
the reverse stroke would be equal. This was
found not to be the case because the downward

Figure 39. Diagram of dashpot arming device.
This device gives longer arming times with larger
values of setback.

stroke was extremely rapid, with the flow prob-
ably turbulent, while the reverse stroke was
slow with laminar flow.

The method of making the electric connec-
tions to the T-132 detonator was another depar-
ture from the practice followed previously.
Contact springs were completely eliminated.
One of the leads of the detonator was grounded
through the rotor driving mechanism, while the
other lead acted both as a mechanical stop and

SECRET

MECHANICAL DESIGN OF PROXIMITY FUZES

193

as the “live” contact. In this manner, the force
of the rotor driving spring was employed to
insure good contacts at both detonator leads.
The details of the detonator rotor and the
methods of assembling the detonator to the
detonator rotor are described in reference 23.
A photograph from this reference is shown in
Figure 40.

The electronic assembly of the T-132 in-
cluded a technique which, while not new in

Figure 40. Jig for forming and cutting
detonator leads in mortar fuzes. At bottom of
photograph are shown from left to right :
detonator, detonator rotor, detonator rotor with
detonator installed, bottom view of detonator
rotor showing lower face of detonator.

general, was new in the case of proximity fuzes.
It consisted of painting the resistors and con-
densers directly upon a ceramic supporting
member. The antenna insulator was also con-
structed of ceramic material, and many of the
oscillator components were also painted di-
rectly on it. The interconnecting leads and some
of the plates of the capacitors consisted of
silver plating on a surface of these ceramic
members. In the original models of the mortar
fuze built at NBS the electronic components
were held between two thin Bakelite plates.

This general design was copied by the Globe
Union Company in their design of the first
ceramic plates. The two plates can be seen in
Figure 41.

From tests in the field and in the high-speed
centrifuge it soon became apparent that the
weight of the components and the potting com-
pound above these plates was sufficient to break
them when under setback. It was immediately
suggested that it would be better to mount the
ceramic plates vertically. Several different am-
plifier assemblies were tested, and it was found
that a single rectangular plate could support all
the necessary components with a great saving
in space. The test terminals were mounted di-
rectly on one of the edges of this plate, thus
eliminating the terminal block of the previous
constructions. A cutaway view of the T-132
with the vertical plate amplifier is shown in
Figure 42.

The antenna insulator was fastened to the
metal shell of the fuze by soldering it to its
silver-plated surfaces. In the case of the T-171
fuze (Figure 43), the ceramic antenna spacer
was replaced by a mica-filled phenolic antenna
block, which was molded directly onto the base
member and which had around its forward sec-
tion a metal ring, to which the shell of the fuze
proper was fastened. This construction resulted
in an extremely rigid assembly, and the later
models of this fuze as well as of the T-132 suc-
cessfully withstood the jolt test.

With the exception of the snap-out shaft de-
scribed above, the arming systems of the T-171
and the T-132 were identical.

The overall dimensions of the T-132 and the
T-171 were determined as follows. It was de-
cided to make the mortar fuze fit the standard
fuze well. The 2-in. diameter was determined
by the availability of the Zenith generator
which, at the time, was the smallest available.
Many shapes of the nose section were tested in
the NBS wind tunnel, and it appeared that the
flat nose cap resulted in the greatest stability
of flight. This cap was, therefore, the one used
in the initial production.

When it became obvious that the flat cap re-
sulted in too great a loss in range and there
arose the possibility of using tail extensions to
improve the ballistics, several caps with better

SECRET

194

MECHANICAL DESIGN

Figure 41. Assembly and principal components of T-132 fuze. Assembly shown uses “horizontal’
ceramic plates which were replaced in later models of this fuze.

MECHANICAL DESIGN OF PROXIMITY FUZES

195

Figure 42. Cutaway of T-132 fuze. On right
center of photograph may be seen vertically
mounted amplifier used in later models of this
fuze. See Figure 34 for identification of com-
ponents.

lining with the same turbine did not prove
worth while.

Mortar Fuze, T-172

A rather radical departure from the T-132
and T-171 mortar fuzes was the T-172 fuze
developed by the University of Florida. In
order to get a more forward-looking angle of
sensitivity of the antenna, a loop antenna was
employed. This meant that the power supply
and most of the electronic components of the
fuze did not have to be isolated electrically

Figure 43. T-171 trench mortar fuze.

assembly, the overall dimensions of the fuze
exclusive of the loop were equal to those of the
T-132. Placing the power supply in the base of
the fuze permitted the direct coupling of the

from the body of the shell. It also indicated the
desirability of locating the generator in the
base of the fuze so as to permit the use of a
plastic cap at the base of the loop. A central air
duct was provided for the intake, and a series
of round holes near the base of the fuze pro-
vided the exhaust. The fuze is shown in Figure
44. By making an extremely compact electronic

streamlining were designed. The modification
adapted for the T-132 and the T-171 was the
132A cap shown in Figure 43 and in Figure 6
of Chapter 1. Further increases in stream-

SECRE'

196

MECHANICAL DESIGN

generator shaft to the arming mechanism. The
gear train was, therefore, located adjacent to
the detonator, and the whole assembly of the
gear train and detonator was made to revolve
into the armed position at the completion of
400 yd of air travel.

One of the problems encountered in the de-
velopment of this fuze was the difficulty of ob-

Figure 44. T-172 mortar fuze.

taining sufficient power from the turbine at the
lowest air velocities. The best shapes of the
intake were such as to increase the drag of the
fuze. Compromise solutions had to be em-
ployed.61’ 72

A special generator was designed for the
T-172 (see Figure 45) 72 by the Zenith Cor-
poration. It was similar to their six-pole T-50
and T-51 generator but used only three coils.

By an ingenious method of assembly of the
stator, the coils were wound directly on the pole
piece assemblies. Another advantage of the de-
sign was the fact that the pole pieces were sup-
ported against setback by a brass ring. This

Figure 45. Three-coil generator for T-172
mortar fuze.

generator was equipped with precision ball
bearings. The first models of the Zenith T-172
exhibited several weaknesses in the method of
supporting the assembled generator, but these
were soon rectified.

Figure 46. Cutaway of T-2005 rocket fuze.

• I

The end of hostilities prevented any large-
scale production of the T-172.

In all three of the mortar fuzes described

M A

MECHANICAL DESIGN OF PROXIMITY FUZES

197

above, a %6-in. thick brass plate was used as
the interrupter below the electric detonator.

Rocket Fuze, T-2005

The success attained with the mortar fuzes
and the advisability of designing a much more
universal rocket fuze than the T-30 and the
T-2004 led to the development of the T-2005.78
Since this fuze was intended primarily for the
Xavy/California Institute of Technology rock-
ets, its physical outlines were designed accord-

Figube 47. Schematic arrangement of arming
mechanism for T-2005 fuze. Arrangement in safe
or unarmed position is shown on left, in armed
position on the right.

ingly. As can be seen in Figure 46, the base
section extends very little into the fuze well,
and the main body of the fuze is mounted for-
ward of the projectile. The general design is
very similar to that of the T-171. The antenna
insulator is made broader and heavier, both to
give increased strength and to result in better
streamlining.

The generator power supply is practically
identical with that of the mortar fuzes. The
pitch of the turbine blades was, of course, re-
duced to result in a speed of 20,000 to 80,000

rpm for projectile speeds of 800 to 3,200 fps.
It was decided to use precision ball bearings for
this fuze, since problems of setback support
and procurement were much simpler than in
the corresponding cases for the mortar fuze.

The arming system of the fuze, however, re-
quires considerable explanation. It was consid-
ered desirable that the use of any arming wires
or manually removable pins should be unnec-
essary, although their use should be provided
for as optional. The fuze should be capable of
being mounted on a projectile below the wing
of a pursuit ship without any danger of being
armed by the airstream. This naturally re-
quired that the rotating system of the fuze be
held from turning until after the firing of the
rocket. The arming system was to operate
when subjected to an acceleration greater than
10 g and perhaps as high as several thousand g.
This fuze should not arm in less than 300 yd
under any condition. If the burning of the
rocket continues beyond 300 yd, the fuze should
not arm until after the completion of burning.
An SD element had to be provided that would
explode the rocket after approximately 6,000 ft
of air travel. This SD feature had to be optional,
to be inserted or removed in the field. The fuze
was to be capable of passing the jolt test. A
simplified drawing of the main components of
the arming system is shown in Figure 47.

A brief description follows: A weight sup-
ported by a spring and provided with a pin in
its forward end acts to lock the turbine in the
fixed position. This weight is normally free to
move back and forth. When the fuze is fired in
the normal manner, this weight moves back,
permitting the air to drive the turbine, the
shaft of which is coupled to a gear train that
lifts an arming rod in a manner somewhat
similar to that in the T-132. This gear train
also performs another function, that of locking
down the setback wreight at the end of 100 turns
of the turbine. This means that if the 10-#
acceleration is maintained for a distance of
roughly 100 yd, the turbine is permanently re-
leased and can continue to operate and then
arm the fuze at the end of 300 yd. For high-^
rockets, and possibly artillery shells, on which
this fuze may be used, another weight was pro-
vided that was retained in its normal position

198

MECHANICAL DESIGN

by a 100-g spring. This high-# weight was in-
terlocked with the low-# weight described above
in such a manner that if the fuze experienced
an acceleration of over 100#, the low-# weight
would move back first, permitting the high-#
weight to move and lock it in the lower position.
This action is identical with that of all the other
double-element setback devices mentioned pre-
viously in this report.

The 300-yd minimum arming distance is not
affected by this action, so that in all cases the
T-2005 fuze cannot arm until a safe distance
away from the launcher. For those cases in
which the rocket burning continues beyond the
300-yd mark, a mercury switch is provided
which keeps the detonator circuit open until the
forward acceleration ceases. An electric RC
increases the safe air travel still further.

Since no double-element setback release was
provided directly in the detonator rotor, the
danger of its arming due to the breakage of
the antenna insulator still existed as it did in
the case of the original mortar fuzes. Because
of the low values of acceleration, a small double-
element safety was not practicable in the det-
onator rotor. A different mechanism was, there-
fore, evolved. The arming shaft, instead of
serving merely as a pin to lock the detonator
in position, was modified so as to rotate as it
was withdrawn. This rotation was communi-
cated to a local arming screw, which also served
to lock the detonator rotor in the safe position ;
that is, the arming shaft was used both as a
lock and as a screw driver. If, while the fuze
was in the safe condition, the antenna insula-
tor was broken and the arming shaft fully
withdrawn, the locking screw would still re-
main in its safe condition, and the detonator
would not move.

The SD was accomplished by a rather simple
form of differential screw. Two small gears of
12 and 13 teeth were arranged to mesh with one
of the pinions of the regular arming gear train.
These two gears were mounted on a fine screw
used as their shaft. One of the gears was cou-
pled to the screw by means of a spline, and the
other was threaded onto it. In this manner, as
the two gears revolved slowly at slightly differ-
ent speeds because of their one-tooth difference,
the screw was slowly moved downward so as to

“ground” the firing circuit at the end of several
thousand feet of air travel. If the use of SD
was not desired, it was possible to disengage
the small gears by the simple removal of a
small screw projecting through the case. The
whole SD assembly was mounted on a small
spring, which normally kept it out of engage-
ment with the driving pinion.

4'3'° Miscellaneous Experimental Fuzes

As mentioned at the beginning of this chap-
ter, many fuzes were considered and experi-
mented with that never saw production. Of
particular interest were the T-40 and T-43
fuzes, familiarly known as Katrinka, which
were intended for operation on very large
bombs. The intention was to use the fin struc-
ture as part of the antenna loop and to make
use of the bomb body by shunt excitation.55
The fuze itself was to be mounted in a cylinder
approximately 6 in. in diameter and 12 in. long
placed inside the fin structure. The energy was
to be derived from a battery that was to be
either very well insulated so as to maintain its
ground temperature for long periods or that
was to be heated electrically by the plane’s
power supply.77

A small turbine was designed to energize the
arming mechanism. A feature of this mecha-
nism of particular interest was the variable arm-
ing that could be controlled from the plane. The
arming system contained a differential gear,
one side of which was driven by the turbine,
while the other side was connected by means of
a flexible cable to a control in the plane. The
output of the differential gear controlled the po-
sition of the detonator. The flexible cable was
to serve both for setting the fuze and for releas-
ing the arming system in the manner similar
to that of an arming wire. Drawings of the
mechanical system were made and some of the
components were actually built, but the project
was abandoned before a working fuze was con-
structed. Other fuzes, particularly the T-51,
appeared capable of fulfilling the application
for which the T-40 and T-43 were intended.

Another experimental fuze was intended spe-
cifically for air-to-air bombing, particularly for

SECRET

THE MOUNTING OF FUZES INTO MISSILES

199

use with toss-bombing equipment (cf. Division
4, Volume 2, STR). This fuze was given the
designation P-4 771B by Bell Telephone Lab-

Figure 48. Experimental model of generator-
powered fuze for air-to-air bombing.

oratories where the development76 was carried
out. A particular requirement for this fuze was
ability to operate satisfactorily at lower air-
speeds at high altitudes. This resulted in a

larger turbine system for the power-supply
generator. Air was directed to the turbine by
scoop or air duct around the periphery of the
fuze. A photograph of the fuze is shown in Fig-
ure 48. Several models were built which operate
satisfactorily against ground targets. No air-
to-air tests were conducted and the project was
abandoned because of the lower priority given
later in World War II to air-to-air weapons.

4 4 THE MOUNTING OF FUZES

INTO MISSILES

Since the vehicle carrying a radio proximity
fuze often acts as an antenna and in all cases
has an effect on the radiation, the method of
fastening the fuze to the projectile is more
critical than with contact fuzes.

For contact fuzes two practices are more or
less standard. If the fuze is shipped as part of
the complete round, it is permanently fastened
in place, usually by staking. Where the fuze is
inserted into the vehicle in the field, this is usu-
ally done without any special tools, and hand
tightening is considered sufficient. No lock
washers of any kind are employed.

In the case of proximity fuzes this procedure
is not tolerable. Loose mounting results in both
mechanical and electric noise. Extensive studies
were carried out6 to determine the best methods
of mounting fuzes in missiles to reduce both
mechanical vibration and poor electric contact.

The M-8 rocket and the mortar fuzes were
designed to be wrench tightened. The bomb
fuzes were to be assembled in the field, prefer-
ably after the bombs were in their racks. To
insure good electric connections, lock washers
were specified. Special wrenches were provided
for all the fuzes. Lugs were provided on the
fuze housing (cf. Figures 16, 20, and 26) to
provide anchor points for the wrenches.

The using Services soon complained that the
Shakeproof lock washers supplied for the T-50
fuzes did not permit easy defuzing. Also, there
was noted on the part of the field personnel a
tendency to use the dipoles of the T-51 and the
T-82 as handles in tightening or loosening the
fuze in its well. Accordingly, a more suitable
washer was developed. It was manufactured by

SECRET

200

MECHANICAL DESIGN

the Shakeproof Company and in its outline was
a duplicate of their external-toothed lock
washer. Instead of twisting the teeth, however,
they were bent alternately back and forward.
This resulted in a spring washer that provided
considerable friction between the bomb and
the fuze without a positive lock action. It also
permitted the operator greater leeway in the
angular setting of the fuze with respect to the
bomb for greater convenience in the use of the
arming wires. Photographs of both the spring

Figure 49. Washers used for mounting fuzes
in bombs. Lock washer is shown on the left,
nonlocking spring washer on the right.

washer and lock washer are shown in Fig-
ure 49.

It was found that the dipoles of the T-51 and
T-82 were sufficiently strong to be used as
handles, when this spring washer was em-
ployed.

45 SPEED REGULATION

Very little has been said so far about the de-
sirability or practicability of automatic speed
regulation for the various windmills and tur-
bines used in the Division 4 fuzes.

There are several reasons why a constant
generator speed is desirable.

1. If the generator could be operated at con-
stant speed at the various velocities of the
vehicle, electric voltage regulation would not be
necessary.

2. The bearing life could be greatly pro-
longed.

3. The vibrational forces could be main-
tained at a minimum, thus increasing the
amount of permissible unbalance in the rotat-
ing systems.

4. The strength requirements of the rotating
system could be eased, permitting a greater

choice of materials for the construction of the
turbine and the rotor assembly.

5. It would permit the standardization of the
drive for the fuzes required for different appli-
cations.

6. By maintaining a reasonably constant
speed, the arming system would be in effect a
time mechanism rather than an air-travel
mechanism. This would be of particular advan-
tage because most of the ballistic tables of the
Services are expressed in units of time.

In contrast to these obvious advantages there
is the disadvantage of greater complexity in-
troduced by the regulating mechanism. This is
particularly serious, since it is very important
that the fuzes be capable of withstanding ex-
treme conditions of temperature and accelera-
tion.

It is, of course, obvious that a great many
mechanisms can be designed to control the
speed of an air turbine. Yet, at the time of the
work, no speed control system was known that
did not require the addition of some moving
parts and that would depend merely on the
aerodynamics involved. The matter was dis-
cussed without success with representatives of
many companies, who normally engage in the
construction of pumps and turbine. In the
summer of 1945, the Douglas Aircraft Com-
pany of California and the University of Cali-
fornia engaged in research on the shape of
nozzles in order to obtain speed regulation for
the T-172 fuzes. (This work was sponsored by
the Transitions Office of NDRC.) Their re-
port00 concludes that some regulation is pos-
sible by designing the nozzle in such a way as
to obtain sonic speeds through the throat. This
will not give perfect speed regulation since the
pressure in the throat is roughly proportional
to the pressure at the nose of the projectile, and,
therefore, the total mass of the air moving
through the fuze will vary with the velocity.
Tests at the National Bureau of Standards on
a similar mechanism of controlling airflow did
not give encouraging results. Nevertheless,
some rather simple mechanisms for controlling
the speed of propellers were tested, and some
bomb fuzes equipped with these were actually
dropped. Perhaps the simplest of these, shown
in Figure 50, was the mounting of the windmill

SECRET

DYNAMIC BALANCING

201

on its hub so that it had some axial freedom.
A spring inside the propeller hub was com-
pressed when the windmill was driven back
against the flat nose of the fuze. This effectively
decreased the airflow and decreased the wind-

Figure 50. Windmill for speed regulation.
Normal position of windmill on its shaft (top) ;
windmill depressed by air pressure (bottom).

mill speed. The speed regulation, while not per-
fect, was quite stable.1’ - The disadvantage of
the scheme lay in the fact that the windmill
had to be somewhat free upon its shaft, result-
ing in a relatively large amount of vibration

and preventing the possibility of good balanc-
ing.

The use of flexible blades in the turbines has
already been mentioned in connection with the
T-82. Another suggestion to use centrifugal
regulation came from Zenith and was investi-
gated by the University of California.60

In both the above schemes flexible members
are employed. These are necessarily located in
the airstream. Consequently, vibrations of high
frequency and high amplitude are set up in the
flexible members, resulting in rapid fatigue.
Another objection to the flexible blade schemes
is the difficulty of maintaining accurate dy-
namic balance at all speeds.

The overall solution of the high-speed prob-
lem was the use of well-balanced rotating sys-
tems, materials of sufficient strength, and the
proper bearings to permit the rotational sys-
tem to withstand high speeds without ill effects.
This method of attack is satisfactory as long
as the maximum velocity range of the projectile
is not greater than approximately 4 to 1. As the
range of velocity of the projectiles equipped
with similar fuzes is increased much beyond
that, speed regulation will undoubtedly have to
be employed.

46 DYNAMIC BALANCING

Although the problem of balancing the ro-
tating system of a fuze may appear to be a pro-
duction problem, not particularly related to
fuze development, appreciable work was done
on the subject by Division 4. While commercial
equipment for dynamic balancing was available
during World War II, such equipment was not
available on the scale necessary for the pro-
duction envisioned. The commercial equipment
was both complicated and expensive and could
not be duplicated by the fuze manufacturers
themselves. As the fuze program advanced, it
became more and more evident that the me-
chanical design of generator-powered fuzes
could be simpler and fuzes would be more re-
liable if the rotating systems were dynamically
balanced. This required that suitable equipment
be available for doing the balancing in produc-
tion.

SECKE

202

MECHANICAL DESIGN

Soon after the T-50 program was started, it
was found that some units were much noisier
than others due to the large amplitudes of
vibrations caused by the rotating systems. A
process of selection was then applied to the
windmills before their assembly. A simple un-
balance tester was built, consisting of a flexibly
mounted fixture coupled to a crystal pickup,

Figure 51. Equipment for dynamic balancing
of vanes of T-50 type fuzes.

which was fed into an amplifier, the output of
which was read on a suitable meter. It was soon
found that it was difficult to distinguish be-
tween the rotational vibration of the fuze head
and the noise due to the rather crude ball bear-
ings employed. A rather sharply tuned filter
was then introduced into the amplifier, and the
speed of the windmill was manually adjusted so
that the rotational vibration was kept at the
frequency at the center of the amplifier peak.
In this way, badly unbalanced windmills were
isolated from the rest.

It was a simple matter to go from this step
to a stroboscope, which was triggered by the
unbalance voltage and indicated the position of
unbalance. The circuit was so arranged that, as
the instantaneous unbalance voltage passed
through zero, it triggered a thyratron which, in
turn, flashed the stroboscope light. The wind-

mill appears to stand still under this light. By
taking a vane and deliberately unbalancing it,
the equipment can be easily calibrated. A pho-
tograph of this equipment is shown in Figure
51. In this method of balancing no effort was
made to achieve true dynamic balance, but
since the windmill can be considered to be a
nearly flat disk mounted at the front end of a
rather large mass hinged at its base, the re-
moval of static unbalance in the vane reduces
the vibration of the large mass to a very low
figure. In the T-50 production no effort was
made to balance the rotor of the generator.

Figure 52. Close-up view of dynamic balancing
machine for rotating systems of mortar fuzes.

The equipment described above was em-
ployed on a large scale by the manufacturers
engaged in the T-50 fuze program.

True dynamic balancing was a requisite in
the construction of the mortar fuzes, and the
machines of various manufacturers were in-
spected for suitability. Since the original inten-
tion in this program was to use ball bearings,
it was important that the dynamic balancing
equipment should be able to distinguish be-
tween ball bearing noise and rotational un-
balance.

It was known that the Westinghouse Com-

CHOICE OF PLASTICS FOR THE PROXIMITY FUZES

203

pany had developed a system of “Micro-
Dynetric” balancing capable of accomplishing
this. A group of NBS and Bowen engineers
visited the Baltimore plant of this company and
witnessed the operation of the only model of
that machine in existence. The machine ap-
peared suitable for the purpose, and orders
were placed by Bowen, Globe Union, and others
for the procurement of this equipment. It be-
came immediately apparent that its production

anced is belt driven at 95 rps. The output of the
amplifier is fed into a vacuum-tube voltmeter
by means of which the magnitude of unbalance
can be determined. The output voltage also
triggers a stroboscope which locates the posi-
tion of unbalance. In a later model automatic
volume control was employed so that no manual
changes of amplifier gain had to be used for a
wide range of rotor unbalance. The removal of
metal was done by hand.

Figure 53. Dynamic balancing machine shown in Figure 52 and accessory equipment.

schedule was very much slower than required
for the fuze program, and NBS undertook to
design a simple and easily produced balancing
machine for the project.9 The machine, photo-
graphs of which are shown in Figures 52 and
53, is similar in operation to the standard ma-
chines of the Gisholt Company and others ex-
cept for considerable simplification. The rotor
drive consists of a synchronous motor so as to
maintain a constant speed. The pickups are two
standard 2-in. permanent magnet dynamic
speakers. The amplifier is very sharply peaked
at approximately 95 c, and the rotor to be bal-

Modifications in this equipment were made
by Raymond Engineering,57 Zenith Corpora-
tion,72 and the Bowen Company,46 for the
mortar fuze production program. Unbalances
of the order of 0.05 g-in. could be detected
readily and corrected.

True automatic balancing in the sense that
the balancing machine either adds or removes
mass in the proper places in the rotating assem-
bly was, of course, considered, but the pres-
sure of work and the termination of World War
II forestalled any work on the several schemes
suggested.

SECRET

204

MECHANICAL DESIGN

4 7 CHOICE OF PLASTICS FOR THE
PROXIMITY FUZES

T-5 and T-50 Type Fuzes

One of the problems presented in the con-
struction of the rocket and bomb proximity
fuzes was that of determining and developing
the proper plastic materials to be used in the
nonmetallic portions of the fuze.

Some of the basic requirements for suitable
plastics were high compression strength, high
impact strength, dimensional stability, and
good electrical properties at high frequencies.
The electrical properties included low dielectric
constant, low power factor, and high leakage
resistance. The main problem, then, was the
search for plastic materials with the above
properties that were commercially available or
could be produced from materials available in
large quantities.

The principal plastic parts of the fuze were
the insulator nosepiece, to be made of plastic
material with good electrical properties com-
bined with good mechanical properties, and the
oscillator block, for which a plastic with good
electrical properties was needed. Since the block
was mounted on a steel plate, high mechanical
strength was of secondary importance. On the
other hand, the terminal plate that seals off the
audio portion of the fuze requires high me-
chanical strength together with high d-c leak-
age resistance. The rectifier housing, the det-
onator rotor housing, and the detonator rotor
all require a plastic material of high impact
strength and good dimensional stability. For
these pieces the electrical requirements are of
secondary importance.

In the original choice of material for the
nosepiece, the electrical requirements had to be
subordinate to the mechanical requirements be-
cause of the particular mechanical design
chosen. There was sufficient space between the
antenna insert and the body of the fuze to cause
the electric gradients to be low enough not to
interfere seriously with the sensitivity of the
unit. Because of this large space, the effects
of humidity on the plastic were also of minor
consideration. Moreover, the high surface

polish of the molded plastic further reduced the
effects of humidity by forming a good surface
seal.

Dimensional stability, absence of cold flow,
and high-temperature heat distortion were con-
sidered points of primary importance in the
design of the nosepiece. Any looseness would
be extremely objectionable when the fuze vi-
brated in operation. Electric noise, which
would result, would produce a spurious signal,
causing the fuze to malfunction. One of the
methods used to fasten the nose to the main
body of the fuze was the use of knurled steel
inserts. The use of these knurled inserts pre-
cluded the use of most of the thermoplastic
materials, not only because of cold flow, but also
because the points on the knurling would cause
stresses which, in turn, would produce crazing
and so destroy the mechanical strength of the
piece. When the nosepiece was further modified
to include the holding of the nose to the base of
the fuze by through screws, the large compres-
sional forces under the screw heads were still
a source of trouble because of crazing.

Tests of various thermoplastic materials,
such as methyl methacrylate and styrene, for
cold flow and crazing, showed that these meth-
ods were not satisfactory for use with the par-
ticular design involved. It was determined that
mica-filled phenolic was the best available ma-
terial. Various brands of low-loss mica-filled
phenolic were tested for compression strength,
creep, electric resistance, dielectric constant,
and power factor. During these tests it was
found that the effective impedance of the mica-
filled phenolic available from the several differ-
ent manufacturers varied by a factor of as
much as 2 to 1, although all the material tested
was submitted as conforming to the same set of
specifications.

In the construction of the oscillator block for
the first fuzes, the same material was used as in
the nosepiece, because these fuzes (oscillator di-
ode type [OD] ) had a tuning condenser molded
into the block. In order to maintain the con-
stancy of tuning, extremely good dimensional
stability was required along with freedom from
effects of humidity. The dimensional stability
was satisfied by the mica-filled phenolic chosen,
and the freedom from effects of humidity was

SE<

CHOICE OF PLASTICS FOR THE PROXIMITY FUZES

205

obtained by finishing the surface of the plastic
properly. For the more recent fuzes, in which
the tuning condenser was eliminated, a styrene
block was used because of its superior electrical
characteristics and the ability of cement to
form a better bond with the styrene. The block
was anchored in place to a metal base plate,
thus giving the sufficient mechanical strength.

In the terminal plate a linen-filled phenolic
was used for its high mechanical strength in a
thin sheet. In order to preserve high leakage
resistance, this plate was boiled in wax to seal
it against the effects of humidity. The main
function of this terminal plate was to hold the
components and potting material in the audio-
frequency portion of the unit. The high leakage
resistance was made necessary by the fact that
one of the test leads was part of a circuit, the
operation of which was affected by a leakage
resistance of 100 megohms.

In the mechanical section of the fuze, which
included the generator housing, the gear train
housing, the rectifier housing, the detonator
rotor housing, and the detonator rotor, a high-
impact phenolic material was used. High im-
pact strength and dimensional stability were
the basic requirements. It was later found that
the dimensional stability of the material was
not sufficient to take care of the small toler-
ances required in the generator housing, and
it was necessary to substitute a pressed metal
housing. Some difficulty was also experienced
with the detonator rotor. Consequently, the
batches which did not maintain their tolerance
because of poor molding were rejected.

472 T-51 Fuze

To modify the T-50 fuze so that the radiation
pattern would appear directly in front of the
fuze, it was necessary to use a transverse bar,
or dipole. In order to conserve developing time
and reduce as far as possible the need for de-
signing new components for this fuze, it was
decided to use as many parts of the original
bomb fuze as possible. To get sufficient mechan-
ical strength with this design and to keep the
overall length of the fuze the same, it was nec-
essary to mount the dipoles at a much lower

point on the nosepiece; consequently, the field
intensity or voltage gradient between the di-
poles and the metal base of the fuze was much
higher. Because of these high field intensities
and because of the desire to obtain greatly in-
creased sensitivities, it was necessary to obtain
an insulating material with the best possible
electrical characteristics. It was already known
from previous experience with styrene and
methyl methacrylate that the heat distortion
point was too low and the material was subject
to cold flow and crazing. However, because of
the need for the superior electrical properties
found only in styrene, a search for a suitable
modified styrene was made and Monsanto’s
Styramic 18 was chosen initially. While its
mechanical strength was lower to a consider-
able degree than that of the mica-filled phenolic,
it was felt that it was still sufficient for the
purpose. Later experience proved that it did
not have quite the mechanical strength desired,
and a modification of this material, Monsanto’s
Styramic 18A, was then used. This material has
proved to be satisfactory in all respects.

Later in the development other material
appeared on the market in quantities sufficient
for production needs. One of the outstanding
materials was Dow Q247, which had all the
desirable mechanical properties of the Mon-
santo Styramic 18A as well as slightly better
electrical properties.

4 7-3 T-171 and T-132 Fuzes

In the T-171 fuze for the 81-mm mortar shell,
Dow Q247 was used. The higher radiation re-
sistance of these missiles and the shorter elec-
tric leakage path (because of reduced size)
made insulating properties of the plastic a
prime consideration. At the beginning of this
project mica-filled phenolics were tried, but the
sensitivity of the unit was only marginal. In
this type of fuze the plastic material had to
support the weight of almost the entire fuze.
The generator as well as most of the mechanical
parts had been moved to the nose of the fuze
to act as an antenna, and the plastic was used
to mount the oscillator and act as an insulating
spacer. Because of improved design of the in-

206

MECHANICAL DESIGN

serts, the problem of cold flow and crazing no
longer had its original importance. In the T-132
fuze no plastic material at all was used, the in-
sulating material being a low-loss ceramic.

474 T-2005 Fuze

Because of its success in the T-51 bar-type
fuze, the same plastic material, namely Sty-
ramic 18A, was used in a fuze designed for the
Navy rockets. The design was somewhat simi-
lar to that of the mortar fuze. The generator
was in the nose portion, and the oscillator was
built in the leading edge of the plastic antenna
insulator. If the development of this fuze had
continued further, possibly a change to Dow
Q247 would have been desirable, as this ma-
terial has much higher flexural strength.

All the components molded of Styramic 18,
Styramic 18A, or Dow Q247 used in the fuzes
described above were quite thick and irregular
compared with the usual piece commercially
molded. Because of the poor heat conductivity
common to all plastics and the thickness of the
sections involved, it was necessary to anneal
the pieces in order to obtain a strain-free prod-
uct. This was accomplished by placing each
complete piece in a tank of hot water and mov-
ing it into successively cooler tanks as the cool-
ing took place. The gradual cooling effect thus
obtained removed internal strains and resulted
in the production of uniformly strong pieces.
In order to obtain moldings of sufficient
strength and density, it was necessary to heat
the plastic almost to the burning point. The
molds themselves were run warmer than in
usual commercial practice. This was done in
order to prevent too thin a case hardening. All
the inserts, dipoles, and nose bearings had to
be preheated in order to prevent too sudden
cooling as the warm plastic reached those points
in the mold.

47,5 Cements

It was desirable in all cases to eliminate as
far as possible the electric noise produced by
mechanical vibration; therefore, the r-f com-

ponents used in the proximity fuze had to be
so firmly anchored together that they could not
be loosened by vibration, temperature variation
during storage, or any shock experienced by
the bombs or rockets.

In order to anchor the different components
in place successfully, it was necessary to use an
adhesive possessing certain qualities: low elec-
tric losses and the elimination of strains and
lift due to the difference in coefficient of expan-
sion between the cement and the component to
which it would be attached. The most fre-
quently encountered base material on which
the cement was to be used was mica-filled
phenolic.

The styrene solutions which were available
during the first stages of production presented
a major disadvantage in their inability to re-
lease solvents readily. Under infrared heaters
24 hours were generally required. If the sol-
vents were still present, the maximum adhesive
strength could not be obtained, and the sol-
vents themselves produced electric loss. When
these styrene solutions did finally become com-
pletely free of solvents, they became so brittle
that they would lift upon the slightest shock.
The latter problem was solved by sand-blasting
of the base material and, consequently, the
roughened surface of the mica-filled phenolic
blocks held the cement mechanically. However,
the roughened surface, in opening the pores of
the material, allowed moisture to be absorbed
much more readily than with the original
smooth hard surface. Subsequently, greater
electric loss resulted. Tests were then made
using a phenolic sirup with powdered mica
added, which was developed by Globe Union,
in order to obtain a final coefficient of expan-
sion of the polymerized mica-filled material
equivalent to that of the mica-filled phenolic
block.

Another approach to the problem was made
by using a mixture of styrene, polymer, and
monomer. In this mixture no solvents were re-
quired to be released, since the monomer poly-
merized to a solid. It was necessary to add
polymer to the monomer for two reasons: (1)
it increased the viscosity of the solution to the
point where it would stay in place, and (2) it
would decrease the shrinkage on polymerizing

•SECRET

CHOICE OF PLASTICS FOR THE PROXIMITY FUZES

207

by the amount of polymer in the solution. When
modified styrene compounds were used for
molding the electric sections of the bar-type
fuze, the styrene adhesive worked especially
well. The adhesive had the same coefficient of
expansion as the modified styrene blocks them-
selves; furthermore, the solvent used formed
one uniform material. However, the inability
of the styrene to release solvents readily caused
the solvent to penetrate the block itself. The
solvent-release problem was thus even more
serious than it had been with the use of mica-
filled phenolic blocks.

The next advance in the solution of the sol-
vent-release problem was the substitution of
plasticized vinyl carbazole, produced by Gen-
eral Aniline and Film Corporation, for the
styrene type of adhesive. The solvent release
time was reduced from about 24 to about 4
hours. This material had almost as good elec-
trical properties as the styrene and satisfactory
mechanical strength. Moreover, because it dis-
solved in the modified styrene base, its coeffi-
cient of expansion was not important.

Dichlorostyrene polymer monomer mixtures
were also found to be satisfactory because of
the compatibility in their use with the styrene
block. One of their important advantages was
the capacity to form an extremely hard glass-
like material, with superior electrical proper-
ties, in about 2 hours. There was, then, no need
for the release of any solvents. One of the dis-
advantages of the mixtures was the tendency
for large forces to be set up upon shrinking in
polymerizing. This difficulty was eliminated by
proper plasticizing.

4/76 Potting

In a further effort to protect the electric
components against the effects of temperature
and humidity and to prevent the production of
electric noise in the amplifier portion of the
units, potting material was poured in place.

The material used had to possess certain
characteristics. It had to have a reasonable
amount of elasticity over an extremely wide
temperature range, low electric losses, dimen-
sional stability, ease of handling, short poly-

merization time, and nontoxic qualities. The
initial material tried was wax. This was un-
satisfactory because of its low melting point
and its tendency to sweat at high tempera-
tures. In order to allow the wax to flow around
the components readily, a high pouring tem-
perature was necessary. Consequently, there
was a tendency for the electrical characteristics
of some of the components being potted to be
altered. At low temperatures the forces pro-
duced by shrinkage were actually sufficient to
fracture some of the glass components.

Because of the trouble encountered with
wax, various addition agents were tried. Two
different mixtures were finally used. One, de-
veloped by Zenith Radio Corporation, consisted
of 80 per cent microcrystalline wax and 20 per
cent polyisobutylene, molecular weight 100,000.
This material was used to hold the oscillator
tube solidly in its tube well. This mixture was
not too brittle at —40 C and did not flow or
sweat materially at -{-60 C.

Another material was a mixture of 20 per
cent ethyl cellulose, 20 per cent beeswax, and
60 per cent ceresin. This material did not have
electrical properties quite so good as the previ-
ous mixture but was much stronger and had
good temperature characteristics. It was found
useful in the centrifugal potting of the genera-
tor stators in the T-171 fuze.

The material used largely throughout pro-
duction was polymerized tung oil. This material
was not entirely satisfactory. It could be poured
into the cavities at room temperature and
would jell in about a half hour. On jelling, it
became a firm rubberlike mass similar to art
gum. It had sufficient elasticity to withstand
shock, and it was firm enough to hold the com-
ponents in place. It was also sufficiently friable,
so that it could be broken up with a knife for
inspection or repair of the units. This material
was thermosetting. Once set, it would not melt
at any temperature, and the shrinkage was
almost nil.

The polymerized tung oil, however, had
rather poor electrical characteristics and was
corrosive toward some of the metallic parts.
Because of this, the electric components were
usually coated with a thin coat of wax before
potting. The speed with which the tung oil set

SECRET

208

MECHANICAL DESIGN

up depended to a certain extent on the amount
of moisture present in it. By eliminating this
moisture, the tung oil set up with much greater
rapidity. The main advantage of eliminating
the water from the tung oil was the increase
in the d-c leakage resistance by a factor of 20
and the decrease in the power factor by a
large amount. The dielectric constant was also
reduced slightly.

Because of the need for a material with bet-
ter electrical and mechanical properties, inves-
tigations were conducted for the development
of insoluble soaps, such as Glidden PT1 and
PT2, that could be poured into a unit in the
liquid state. Saponification would thus occur in
situ. When these soaps were substituted for
tung oil, the incidence of certain types of re-
jects in production was changed from a normal
11 to 1 per cent. This material, however, had
disadvantages. Its viscosity, which was higher
than that of tung oil, somewhat hindered pour-
ing. Its water resistance was not so good as
that of tung oil, however, because the material
was in a closed space. This had no ill effects on
the operation or storage of the units. Also, the
material did not have as high a mechanical
strength as tung oil although it was found ade-
quate for the purpose.

Because both of the above materials left
something to be desired both from the electrical
and the mechanical standpoint, work was done
on the use of styrene co-polymers; Dow Q344
and Dow Q349 are probably the samples of the
best available material of this kind. All the
electrical and mechanical properties of the final
set of these materials were completely satisfac-
tory. The initial viscosity made them somewhat
difficult to handle. Furthermore, the surface
had to be sealed from the air to eliminate
stickiness as air hindered the surface polymeri-
zation. These materials were used by the Wur-

litzer Company for potting the oscillator com-
ponents of the T-171.

4 7 7 Solder Flux

In the examination of a number of units over
a period of time, units which were maintained
especially for aging tests, it was found that
some of the electric measurements in some of
them were subject to a constant drift. When
these units were opened and examined, corro-
sion was discovered around some of the sol-
dered joints. At first this corrosion was thought
tp be due to the corrosive action of the tung oil
potting material, until it was realized that the
soldered joints were protected from the tung
oil by a thin wax filament. This corrosion was
then subjected to a chemical analysis and found
to be a metal resinate. The resin which formed
this resinate could have come only from the
rosin in the solder flux.

Because of the almost impossible task of re-
moving all of the flux from the finished soldered
joint, it was desirable to investigate other
fluxes which were thought to be less corrosive.
The corrosion was produced by the absorbed
oxygen, as the soldered joints were completely
sealed by wax and tung oil from the air. In
order to correct this situation, several other
materials were examined as to their suitability
for solder flux. One of the materials examined
is known as polypale rosin, which is a rosin
dimer. This material absorbs only half as much
oxygen as is absorbed by ordinary rosin, and
the corrosive effects are cut down proportion-
ally. Upon testing, it was found that polypale
rosin was also considerably superior to ordi-
nary rosins as a flux because of its superior
wetting qualities. This material has been used
in the fuze production with excellent results.

t SECRET

Chapter 5

CATALOGUE OF FUZE TYPES'

51 INTRODUCTION

General Remarks

IN THE PRECEDING CHAPTERS of this Volume,
there were discussed the general military
requirements of proximity fuzes, the basic
theory of operation of radio proximity fuzes,
and the fundamental principles of design of the
important parts. The requirements of ideal
fuzes were defined, and the limitations that
are imposed by fundamental considerations
were discussed. It was made clear that a combi-
nation of fundamental and practical factors
made it necessary to design different fuzes for
different purposes. It was shown that the de-
sign of a fuze was affected by complex prob-
lems of availability of components and by the
need to make use of facilities and subassem-
blies provided by the development and produc-
tion of fuzes of earlier design. Before launch-
ing upon a discussion of the manifold problems
of producing the fuzes in large quantity and
testing them in the laboratory and in the field,
it is desirable to present a description of the
various fuzes that were produced. It is the pur-
pose of this chapter to provide such a descrip-
tion. Furthermore, at the expense of some repe-
tition, the chapter may be read separately from
the rest of volume, without undue loss of mean-
ing although frequent reference is made to fig-
ures elsewhere in the other chapters.

The description that is given in this chapter
is intended to provide for each fuze (1) a state-
ment of the principal applications for which
the fuze was designed and the limits, so far as
known, within which satisfactory performance
may be expected, (2) performance characteris-
tics under typical conditions, (3) engineering
data that are useful in the estimation of per-
formance under certain conditions which are

a This chapter was prepared by Thomas N. White, Jr.,
with the assistance of Rachel Vorkink, Paul F. Bar-
tunek, Alan L. Leiner, and Rosalind Schwartz, of the
Ordnance Development Division of the National Bureau
of Standards. Bartunek is now with the Physics Depart-
ment at Lehigh University.

not covered herein, (4) miscellaneous data on
important characteristics that distinguish one
fuze from another, and (5) summary data
charts for each fuze. For a full understanding
of the terms used in the development of items
(3) and (4) above, some reference to other
chapters may be desirable. Such references are
indicated in the following presentation.

Sources of Data and Acknowledgments. The
scope of the chapter, as outlined above, is some-
what broader than is demanded by the logical
development of this report. Information is pre-
sented that demands substantiation in subse-
quent chapters. This anticipation of results has
the advantage that it permits the orderly pres-
entation, in a single chapter, of the essential
characteristics and limitations of each fuze.
Enough has been said in preceding chapters so
that the data in this chapter can be clearly un-
derstood. Enough will be given in following
chapters to show the variations to which these
data are subject. The discussion of variations
and difficulties of production, and of testing in
the laboratory and in the field, can be under-
stood more readily with reference to the aver-
age properties of the fuzes as actually pro-
duced.

An effort has been made to select data that
are representative of the bulk of production
fuzes. Available data on experimental fuzes are
also included in so far as possible.

It is important to note that many of the data
given in this chapter are average values , or
“best estimates/’ The characteristics of indi-
vidual fuzes differ more or less from the aver-
age. A full account of the individual variations
would lead to undesirable complications in this
chapter. Discussions concerning the occurrence
of individual variations and the reliability of
the estimates are taken up in other chapters of
this report.

It is appropriate at this point to acknowledge
the courtesy of the Ordnance Department and
the Signal Corps in providing much valuable
information on the performance of production
model fuzes in acceptance tests. Much of the

SECRET

209

210

CATALOGUE OF FUZE TYPES

most useful and reliable information on fuze
performance under standard conditions came
from these sources. Acknowledgment is made
to these sources and also to the Army Air
Forces Proving Ground, Eglin Field, Florida,
the Naval Ordnance Proving Ground, Dahlgren,
Virginia, and the Naval Ordnance Test Sta-
tion, Inyokern, California, for special informa-
tion on certain important experimental and
service tests. In order to avoid complications it
has been necessary to omit references to specific
sources of data in this chapter. The reader can
obtain a full appreciation of the value of the in-
formation obtained from military sources only
by a study of other chapters, particularly Chap-
ter 9 of this report.

Scope. The fuzes covered in this chapter are
primarily those which reached large-scale pro-
duction. Data on experimental fuzes are pre-
sented in Chapters 3, 4, and 9 and briefly in
this chapter in Section 5.6.

Where the military requirements for per-
formance are presented in this chapter they
refer generally to production specifications
rather than to the original requirements for the
development project.

The technical specifications recommended by
NDRC to the services for production fuzes are
included in the bibliography.

0,1 2 Developmental Relations between
Fuzesb

The first radio fuze developed was the longi-
tudinally excited battery-powered T-5, designed
for use on the 4.5-in. Army rocket M-8 in plane-
to-plane firing. The T-6 was the same fuze pro-
vided with an extended arming time to make
it suitable for ground-to-ground firing on the
same rocket.

The first bomb fuzes that were produced in
quantity were members of the T-50 group.
These fuzes may be regarded as T-5 fuzes modi-
fied, as required, by the introduction of a wind-
mill-driven generator and arming system and

b The historical aspects of this account are overly
simplified. For a much more detailed and accurate treat-
ment, see reference 2 and the history of Division 4,
NDRC.

provided with a larger antenna and different
oscillation frequencies. This antenna, in the
form of a ring, led to the name ring-type fuze.
In order to make the most of existing produc-
tion facilities, the layout of the radio and audio
circuits of the T-5 were maintained essentially
intact. Two carrier frequencies, Brown for the
T-50-E1 production model and White for the
T-50-E4 production model, were found desir-
able in order to obtain satisfactory burst
heights of bombs in the size range 100 to 1,000
lb. A number of experimental models were
used, of which the most important were the
T-50-E10 (Brown) and T-50-E3 (White).
These were altered from time to time to try out
changes in the radio and audio circuits and for
other purposes. The production fuzes T-89,
T-90, T-91, and T-92 were T-50 type fuzes im-
proved in certain respects and modified to make
them more suitable for certain types of bomb-
ing.

The rebuilding of the T-5 to provide a group
of longitudinally excited bomb fuzes was rec-
ognized at the outset as an expedient, in that
longitudinal excitation led to a considerable
dependence of the burst height of the bomb
upon the conditions of release (altitude and
speed). This variation in performance was re-
duced somewhat through the use of a suitable
amplifier characteristic, but the results repre-
sented some compromise with ideal require-
ments.

Accordingly, at the same time that the ring-
type bomb fuzes were being developed, work
was carried on (on a second priority basis) to
develop a transversely excited (bar-type) c
bomb fuze, the T-51. In order to take full advan-
tage of the benefits of transverse excitation,
considerable effort was made to obtain an am-
plifier characteristic that was relatively flat
throughout the expected range of doppler fre-
quencies. The performance characteristics of
this fuze were markedly superior to those of the
ring-type fuzes in certain important respects,
particularly the relative independence of burst
height on bomb size and release conditions, and

c The terms ring- and bar-type were in use so exten-
sively that they are now retained. The more fundamental
terms, longitudinally and transversely excited, are used
for fuzes that did not actually have rings or bars.

RET \

INTRODUCTION

211

the greater burst heights attainable. Conces-
sions to expediency in the matter of using the
power supply and arming system developed for
the ring-type fuzes made it possible for the
T-51 to overtake the production rate of the
T-50 group within a relatively short time.

A parallel but slower development, with min-
imal concessions to expediency, was that of the
T-82 bar-type fuze. In this development par-
ticular emphasis was placed on the avoidance of
disturbances in the radio- and audio-frequency
systems arising from moving mechanical parts.
In place of the windmill with its shaft running
through the radio and audio block, an air duct
was carried through the block to a base-
mounted turbine. The T-82 fuze had also other
advantageous features, but it had not been
carried into full production at the close of
World War II.

The ring-type bomb fuze, which, as men-
tioned above, had its origin in the fuze for the
4.5-in. Army rocket, was later modified for use
on Navy aircraft rockets [AR] and high-
velocity aircraft rockets [HVAR]. In this case
the principal structural change was the intro-
duction of an arming delay mechanism that
prevented the arming of the fuze until a certain
time after the burning of the rocket propellant.
Two types of fuze were produced, both in the
Brown carrier band. The T-2004 (Mk-172 Mod
0), intended primarily for plane-to-ground (or
water) firing with the 5.0-in. AR rocket, was
in production at a relatively high rate at the
close of World War II. The T-30 (Mk-171
Mod 0) for plane-to-plane firing with the
HVAR, was not so urgently needed and was not
carried into full production.

Certain important developmental relation-
ships are apparent in the structure of the group
of rocket and bomb fuzes discussed above. An-
other group of fuzes, the so-called “miniature”
fuzes, also shows certain close relationships
among the members of the group. This group of
fuzes was developed later, and considerable use
was made of the information obtained during
the development of the larger fuzes, although
the structural relationships are not so appar-
ent. The most advanced member of the minia-
ture group was the longitudinally excited T-132
fuze for trench-mortar shells. One outstanding

characteristic of this fuze, which was rapidly
approaching the production stage at the close
of World War II, was the use of circuit connec-
tions made by painting or spraying material
through a template onto a ceramic block. Other
members of the miniature group that were de-
veloped to a more or less advanced stage were
the T-171 mortar shell fuze (incorporating
standard electric components), the T-172
mortar shell fuze with a loop antenna, and the
T-2005 general-purpose [GP] rocket fuze.

Performance Terminology

Certain important terms used to describe
fuze performance require definitions. These
are :

Proper Function. [Abbrevation : proper or
P.] A fuze function attributable to normal in-
teraction between the fuze and target.

Random Function. A spontaneous fuze func-
tion, or one not attributable to interaction be-
tween the fuze and the target. In Chapter 9,
the random functions are called either “early”
or “middle” functions for reasons that are there
made clear. For the purposes of this chapter,
there is little need for such a distinction and the
term random function, which has been used
extensively in the theaters of operation, is
applied.

Sympathetic Function. The functioning of a
fuze caused by the burst of a neighboring vari-
able-time [VT] fuzed projectile (e.g., in salvo
or train releases). Sympathetic functions may,
under certain conditions, be caused by either
random or proper functions, and in such cases
are called sympathetic random or sympathetic
proper functions, respectively.

Radius of Action [ ROA ]. A measure of the
proximity to an airplane, or like target, within
which reasonably reliable functioning of a VT
fuze can be expected. The ROA is usually de-
fined as the radius of a cylinder, with axis
parallel to the trajectories, within which a
specified percentage of proper functions should
occur.

Afterburning. (In connection with rockets.)
Burning of residual propellant after the end of
the main blast. Afterburning that persists be-

SECR

212

CATALOGUE OF FUZE TYPES

yond the fuze arming period is conducive to
random functioning of the early variety. After-
burning is aggravated by conditions such as
low temperature or inadequate charge that
bring about inefficient combustion of the pro-
pellant. The phenomenon of random function-
ing caused by afterburning is complex and not
yet understood in full detail. For a thorough
discussion of the subject, see Chapter 9.

Preparation of Fuzes for Use

For all fuzes that were produced on a large
scale, Army and Navy manuals are available
in which full details are given on the prepara-
tion of the fuzes for use. For the Army rocket
fuzes see references 5 and 6. These fuzes are
preferably checked shortly before use by means
of field test equipment which is described in
the bulletins. For the ring-type and bar-type
bomb fuzes, instructions are given in refer-
ence 7. For Navy rocket fuzes, see reference 4.
Some of the descriptive material in this chapter
has been taken from these manuals.

Safety and Arming

All VT fuzes have two common safety fea-
tures: (1) an off-line powder train, and (2) an
interrupted electric detonator circuit. It is the
purpose of the arming mechanism to line up the
powder train and to complete the electric det-
onator circuit after the projectile has traveled
to a safe distance. In considering the general
characteristics of the various arming mecha-
nisms, it is convenient to divide the fuzes into
two classes: (1) fuzes for relatively non-
accelerated projectiles (bombs), and (2) fuzes
for accelerated projectiles (rockets, mortar
shells) .

1. All the fuzes for relatively nonaccelerated
projectiles have a windmill-driven generator
and arming mechanism. In these fuzes the
windmill or vane (see footnote h Section 3.4.5)
must be turned a definite number of revolutions
in order to arm the fuze, and also, after arm-
ing, the vane must be turning at a certain mini-
mum rate in order to provide sufficient voltage

to sensitize the fuze. For projectile speeds with-
in a fairly wide range, the rotational speed of
the vane is very nearly proportional to the
speed of the projectile, so that the distance
along the trajectory to arming is practically
independent of the speed of launching. How-
ever, since the ratio of. vane speed to projectile
speed is not the same for all sizes and shapes
of projectiles, the air travel to arming is not the
same for all fuze-projectile combinations.

2. All the fuzes for accelerated projectiles
are so designed that acceleration in the proper
direction is required for arming. These fuzes
will not arm if subjected to an acceleration that
.is substantially less than the minimum to be
expected with the projectiles on which the fuzes
are to be used. Furthermore, it is necessary
that the acceleration should persist for a cer-
tain minimum time, i.e., that the projectile
should attain a certain minimum velocity be-
fore arming can occur. Other arming require-
ments differ for the different types of fuzes,
varying from a fixed arming-time requirement
for Army rocket fuzes to a fixed air-travel re-
quirement for mortar shell fuzes.

All fuzes having an arming vane are
equipped with an arming pin which prevents
the vane from turning until the projectiles are
released.

An additional safety feature used on some
models is the safety pin. The safety pin is in-
serted into the arming mechanism through an
opening in the booster cup. The pin cannot be
inserted unless the arming components are in
the safe unarmed position. Each fuze comes
supplied with this pin in place and the fuze
cannot be inserted into the fuze well unless the
pin is removed. A most important feature of
the safety pin is that fuzes whose seals have
been removed can have the arming mechanism
checked for safety in the field.

The arming features are built into the fuzes
and can be altered only by breaking seals or by
other deliberate action. Although no adjustable
arming mechanism is provided in the fuzes, it
is possible to extend greatly the air travel re-
quired to arm most of the bomb fuzes by the
use of an accessory device called an “arming
delay” (see Figure 1 of Chapter 4). This de-
vice is set for the desired delay distance and

SECRET

FUZES FOR THE ARMY 4.5-IN. ROCKET

213

is clipped onto the bomb fuze. After the set
delay distance has been traversed the delay
device is thrown off, releasing the arming vane.
From that point on arming proceeds in the
usual fashion.

All fuzes are detonated by the discharge of a
condenser through an electric detonator. After
the condenser has been charged, it may remain
charged for some time even if the source of
electric power ceases to function. For this rea-
son, fuzes which are known to have been
armed, such as duds found on the ground,
should not be handled for one hour after the
vanes have stopped rotating. Duds should be
handled only by qualified bomb disposal officers.

5 2 FUZES FOR THE ARMY 4.5-IN. ROCKET
521 General

Military Requirements

The Army M-8 rocket and the VT fuze for it
were developed at about the same time under
conditions of great urgency, primarily as a
means of defense against bombing attacks. The
VT-fuzed rocket was to be fired from fighter
planes against bomber formations. Although
the original requirement prior to development
was for a 50 per cent fuze, it was required for
production items that at least 65 per cent of the
fuzes should function if the rockets passed
within approximately 60 ft of a target plane.
It was required also that the fuze should oper-
ate at a point on the trajectory such that effec-
tive use would be made of the lateral concentra-
tion of fragments from the rocket. Since the
relative velocities of rocket and target, the
angle of attack, and the structure of the target
were all variable, it was not practicable to
specify any sharply localized set of burst posi-
tions. In general, however, it was desired to
have the rocket burst just before it arrived
closest to the target plane (see Section 1.3).

The VT-fuzed rocket was later shown to have
properties that made it adaptable for use in
strafing operations or in ground artillery oper-
ations, but it was not designed for these pur-
poses.

Although it is not within the scope of this re-

port to detail the characteristics of the rocket,
some remarks on this topic are required to per-
mit a balanced assessment of the usefulness of
the fuze. In particular it should be noted that
the rocket was not adapted to precision shoot-
ing. As a result, even with a perfect fuze the
probability of disabling an isolated enemy tar-
get plane would have been appreciably reduced
by the dispersion.

Fuzes and Rockets

The T-5 and T-6 VT fuzes for the 4.5-in.
Army rockets are intended for use as indicated
in Table 1.

Table 1. Application of T-5 and T-6 fuzes.

Fuze

Use

4.5-in. rockets

PD, T-5

Plane to plane

M-8, M-8A3, T-22, T-74,

Plane to ground

(M-8A1, M-8A1B1,

Plane to water

M-8A2)*

PD, T-6

Ground to ground

M-8, M-8A3, T-22,
(M-8A1, M-8A2) f

* Fuze T-5 should be used in these rockets only when the fins have
been notched (see reference 6, paragraph 10), or when modified by
4.5-in. aircraft rocket kit T-23.

t T-6 fuze should be used on these rockets only when the fins have
been notched (see reference 5, paragraph 10).

The fuzes T-5 and T-6 screw directly into all
standard loaded 4.5-in. rockets listed in Table 1.
They are directly interchangeable with the
PD M-4 series of rocket contact fuzes both
physically and ballistically. The fuze as issued
is not complete. A battery must be installed
prior to use. The standard components of a
fuze are shown in Figure 12 of Chapter 4.
Figure 1 shows 4.5-in. rocket M-8 fuzed with
T-5. The fuze T-5 has a 1-sec arming time ob-
tained through use of the switch SW-230A or
SW-230C, 1.0-sec arming. The fuze T-6 arms
in 3 to 6 sec, by using the switch SW-230A or
SW-230C, 5-sec arming. The only other differ-
ence between the T-5 and T-6 is that the T-5
contains a self-destroying feature that will
detonate the rocket approximately 6 to 12 sec
after being fired if the fuze has not already
functioned on a target.

General Limitations

The fuzes may be used during day or night
and are not affected by fog or clouds. Heavy

SECRET

214

CATALOGUE OF FUZE TYPES

rain will increase the number of random func-
tions and duds. On account of the use of dry cell
batteries as a power supply, extreme tempera-
tures must be avoided. Satisfactory operation

Figure 1. T-5 fuze on M-8 rocket.

can be expected in the temperature range from
20 to 100 F.

The fuzed rockets may be fired in ripple salvo
without sympathetic functioning caused by
random bursts.

The requirement for notching of the fins,
mentioned above, is introduced because some of

the M-8 rockets were manufactured with fins
that did not lock in the open position. Any rat-
tling or vibrating electric conductor on the sur-
face of the rocket, which is part of the radiat-
ing system of the fuze, is likely to cause electric
disturbances that will detonate the fuze. In
notching the fins on the rockets that require
this treatment, care must be exercised to avoid
excessive twist of the fins. Otherwise the rock-
ets may be caused to spin at a rate that will
delay the operation of the arming switches (see
following section). This precaution is of con-
siderably greater importance in the use of the
T-5 than in the use of the T-6 fuze.

The T-5 fuze is susceptible to random func-
tioning under conditions that aggravate after-
burning of the rocket propellant. One type of
propellant trap that was used during the de-
velopment of the rocket was found to be con-
ducive to a high incidence of random functions
about 2 sec after firing. This particular trap,
characterized by a double ring of wire at the
rear end, was not used in the final rocket de-
sign.

Functioning Characteristics
Safety, Arming, and Self-Destruction

The arming of the fuzes is controlled by a
mechanism that delays the arming for a certain
period of time after the end of the acceleration
that occurs during the burning of the rocket
propellant. A detailed description of the me-
chanical arming is given in Chapter 4 and of
the added RC arming in Section 3.3. The arm-
ing distance thus depends upon the velocity of
the rocket prior to arming. The velocity of the
rocket is dependent upon a number of factors,
such as weight of the round, the amount and
temperature of the propellant, and the speed of
the plane if the rocket is fired from a plane.
There are also variations from fuze to fuze in
the arming distance, on account of tolerances
permitted in manufacture. The reader is re-
ferred to Chapter 9 for a discussion of these
factors. For the present purpose it is. sufficient
to give as arming distance the minimum dis-
tance at which arming will occur under any
reasonable conditions of firing.

SECRET

FUZES FOR THE ARMY 4.5-IN. ROCKET

215

In determining the minimum range from
which VT-fuzed rockets can be fired profitably,
it is necessary to know the maximum arming
distance, namely that distance at which all the
fuzes will be armed. This distance also depends
upon a number of factors, but a reasonable
upper limit can be given.

The arming switch is so designed that it can-
not be operated by violent jolts, nor can it be
assembled into the fuze unless it is in the safe
position. Although very unlikely, arming is not
impossible in case of rocket blowup, and after
such events fuzes should be disposed of only by
trained personnel. This statement applies also
to dud fuzes.

The self-destruction [SD] feature mentioned
above is incorporated into the T-5 fuze pri-
marily as a safeguard against operation as a
ground-approach fuze when used over friendly
territory in case it does not operate on a target.

The arming mechanism was not designed for
a spinning rocket. As mentioned above, exces-
sive twisting of the fins during the notching
operation required by a few of the rockets can
cause a spin that will interfere with proper
switch operation. Experience has shown that if
the fins are not twisted in excess of 2 degrees
no trouble will be encountered (see Section 9.2
for details) .

For the T-5 fuze, the risk of a random burst
within certain specified distances is indicated
in the following table.

Minimum range (yd) for plane-to-plane use,
both planes the same speed.

Plane

speed

Attack

(mph)

Pursuit

from side

Head-on

300

243

390

537

400

235

430

625

For the T-6 fuze the probability of arming
within certain specified distances is given be-
low. The distances that are of practical interest
are the horizontal distances for minimum
quadrant elevation [QE].

Horizontal distances Probability of
from launcher arming

(yd) (percent)

840 nil

900 1

960 5

The minimum firing range is determined
primarily by the need to use a QE large enough
to carry the just-armed fuze so high that it will
not function on the ground signal. The mini-
mum QE and range are 8 degrees and 1,600
yd. Another reason for placing a lower limit
on the QE is in order to avoid excessive burst
heights (see following section on ground-to-
ground firing of the T-6).

Proximity Bursts in Plane-to-Plane
Firing of the T-5

Distance from
launcher
(yd)

175

210

250

Risk of

random function
(per cent)

nil

In aerial combat the position of the target
plane will change appreciably during the arm-
ing of the fuze so that allowance must be made
for the relative velocity of the attacking and
target planes in estimating minimum firing
ranges. A representative table of minimum fir-
ing ranges follows. At these ranges a negligible
percentage of the fuzes are unarmed. The
values given for broadside approach are appli-
cable in plane-to-ground firing.

The radius of action of the fuze is about 60 ft
for attack on a medium bomber from the rear.
For other forms of attack, it may vary some-
what, in the manner indicated in reference 6,
Figure 2. On the average at least 75 to 85 per
cent of VT-fuzed rockets within ROA can be
expected to yield proximity bursts. The lower
percentage is to be expected at extreme range
on account of greater losses through random
functioning during the longer flight.

The distribution of bursts about an airplane
target is too complex for description here (see
Section 9.2.3). In general it may be stated
that for a round that passes very close to a
part of the target, e.g., the tail, the burst will
occur almost opposite that part of the target.
For rounds that pass close to the ROA, the

SECRET

216

CATALOGUE OF FUZE TYPES

burst is likely to occur about opposite the center
of the target.

Plane-to-Ground Firing of the T-5

The mean burst height varies considerably
depending upon the dive angle (see Figure 31,
Section 5.5). It should be noted that for shallow
dives the bursts are very high and widely scat-
tered, diminishing considerably the accuracy
of placement of the bursts and the damage to
be expected. For best results, dive angles should
be in excess of 30 degrees.

Ground-to-Ground Firing of the T-6

Variation in burst height with firing QE is
shown in Figure 32, of Section 5.5. It may be
seen that, for angles of 30 degrees or less,
heights obtained over ground would center
around 70 ft with a very large scatter. Effect-
field tests have indicated that such a height
would be excessive under most conditions. To
obtain more satisfactory heights, therefore,
firing elevations should be about 40 to 50 de-
grees.

5-2-3 Structure

General Arrangement

The nose member (MC-382) of the fuze,
comprising the electronic system, is outlined
by a hollow conical plastic shell mounted on a
metal platform which supports the r-f oscillator
block inside the shell (cf. Figure 12, Chapter
4) . The amplifier section is potted inside a thick
metal skirt which extends down from the plat-
form. The tip of the conical shell is metallic
and serves as antenna for the fuze.

The battery member (BA-75), outlined by
an insulating container, comprises an axially
located A supply cell nested in a cylindrical
firing condenser which is surrounded by six
stacks of B supply cells and one additional A
supply cell.

The switch member (SW-230C), also cylin-
drical in outline, contains an acceleration-
operated mechanism for closing the A and B
circuits and for delayed closing of the firing
circuit and alignment of the powder train. In
the switch is located the electric detonator, a

tetryl lead, and in some switch models a me-
chanical SD switch.

The nose and switch carry projecting con-
nector pins and the battery, located between,
has corresponding socket holes on each end.
The three members, when plugged together, are
screwed into a housing (M-381) which con-
tains a tetryl booster charge at the bottom.
This then comprises the complete fuze, T-5
or T-6.

Arming Mechanism

The arming process is completely controlled
by the mechanism contained in the switch
SW-230C. The mechanism is based on a new
acceleration-integration principle developed at
the National Bureau of Standards [NBS] and
fully described in Chapter 4. The switch will
not operate unless subjected to an acceleration
greater than 75# in the proper direction for a
time greater than 0.15 sec. (The principal ele-
ments of the switch are shown in Figure 5,
Chapter 4.) In brief, the operation is as fol-
lows : The acceleration, acting on a weight
eccentrically located on the driving shaft of the
switch turns the shaft 90 degrees against the
force of a 75 -g spring. This motion is retarded
by an escapement wheel and flutter-bar (see
left-hand view in Figure 5, Chapter 4). At the
end of the 90-degree turn, a spring-loaded
rotary switch (center view) is allowed to snap
shut (right-hand view), closing the A and B
battery supply circuits of the fuze. When accel-
eration ceases, a rack-toothed slider-bar is
driven by means of a pinion gear over to the
other side of the switch, closing the firing cir-
cuit contacts which are located at the end of
the slider channel, and also aligning a tetryl
plug in the slider with an electric detonator
which lies at the center of the channel. The mo-
tion of the slider is retarded by the same
escapement mechanism to the extent of 0.7 sec
or more. Additional arming is provided in some
of the SW-230C switches by insertion of a
resistor in the thyratron-condenser circuit.
This delays the time at which the condenser
will acquire sufficient energy to fire the detona-
tor. Arming times up to about 6 sec are secured
in this manner. The total arming time is
stamped on the SW-230C switches.

FUZE FOR NAVY ROCKET AR 5.0

217

Self-destruction of the T-5 is accomplished
either by an RC circuit containing a special
neon bulb NE-23 or by a mechanical switch.
The circuit for the RC-SD consists of a 30-
megohm resistor connected from B-f- to a
0.25-mf condenser, the other side of which is
grounded (see Figure 52, Chapter 3). From
the common point of resistor and condenser,
the neon bulb connects to the thyratron grid
feed line at a point between R-15 and R-16. The
mechanical SD is illustrated in Figures 4 and
10, Chapter 4.

R-F System

The oscillator diode [OD] circuit used in the
T-5 and T-6 fuzes is shown in Figure 2. Values
of the components are given in Table 2.

Table 2. Component values for oscillator in T-5
and T-6 fuzes.*

Resistor

(R)

Value

Con-

denser

(C)

Value

(mmf)

Coil Notes
(L)

0.1 megohm

6 turns

15,000 ohm

6 turns

10 ohm

variable

5 turns

0.1 megohm

r-f choke

1.0 megohm

r-f choke

Triode: QF 200 C or SA 780 A
Diode: QF 197

* Switches SI and S2 are not located in the oscillator section, but
in the SW-230 switch section; C18 is located in the amplifier section.

The oscillator components are assembled on
a phenolic block much as illustrated in Figure
5, Chapter 6. The circuital relation of the com-
ponents can perhaps be visualized better from
Figure 13, Chapter 3. Both illustrations are
actually of the T-50, as evidenced by the central
hole for the generator shaft; exact details of
the T-5 oscillator block are given in reference 1.

Amplifier

Except for a few minor elements, the ampli-
fier circuit of the T-5 and T-6 fuzes is shown
in Figure 25, Chapter 3, to which Table 3
applies. The missing elements are noted in the
table.

Table 3. Component values for amplifier in T-5
and T-6 fuzes.

Resistor

(R)

Value

(megohm)

Condenser

(C)

Value

(mf)

1.0

0.02

0.15

50 mmf*

1.0

0.001

3.3

250 mmf

0.68

0.001

1.0

0.002

1.0

0.01

1.0

3.3

Coil

Turns

2.2

(L)

0.1

4.7

19 in. of No. 32

advance wire
wound on C9
(resistance and
inductance
shown in Fig-
ure 9)

Pentode: QF 206 or SA 781 A.f
Thyratron: GL 489 (GE) or SA 782 B.y

* The following three 50-mmf by-pass condensers do not appear on
Figure 9, Chapter 3; pentode grid, pentode filament, thyratron grid
to ground.

t For SA 782 B, make R17 = 0 and R19 = 2.0 ohm.

X For SA 781 A, make R15 = 1.0 megohm, R9 = 4.7 megohm, and
C14 = 0.005 mf.

5 3 FUZE FOR NAVY ROCKET AR 5.0

531 General

Military Requirements

The VT fuze designed for plane-to-surface
application of the AR 5.0 rocket was required
to give the following performance: (1) proper
function scores should be on the average
greater than 70 per cent, and (2) burst heights
should lie within the range 10 to 100 ft.

Fuze and Rocket

Designations for fuze and rocket parts are as
follows :

Fuze

Rocket

Motor

Head

T-2004 Mk-172

AR 5.0

3.25-in.

5-in.

ModO

MK-7

MK-1

(Army (Navy

designa- designa-
tion) tion)

The fuze is a modified ring-type bomb fuze
and in external appearance is identical with

SECR

218

CATALOGUE OF FUZE TYPES

the T-50 models. (It requires a larger cavity
than the MK-149 or other rocket nose fuzes
and is, therefore, not interchangeable with
them.) See Figure 3 for complete round, T-2004
on AR 5.0. The principal structural change was
the substitution of a setback gear train in place
of the bomb fuze gear train. The function of
this gear train is to complete mechanical arm-
ing at the end of the burning period. A second

energy to fire the detonator. In firing at short
range, this can cause duds.

Functioning Characteristics
Safety and Arming

The mechanical arming mechanism is de-
scribed fully in Chapter 4. Additional time de-

Figure 2. Oscillator-diode circuit used in T-5 and T-6 Army rocket fuzes.

change was the introduction of a delay that
prevented arming of the fuze until a certain
time after burning of the propellant had
ceased.

General Limitations

The fuzes may be used at any temperature
at which the rocket can be used. They are not
affected by clouds, fog, snow, or light rain but
may be affected by heavy rains or hail.

Afterburning of the rocket propellant may
cause early functions if the fuzes are armed.
Another effect of serious afterburning is to
cause repeated dumping (see Section 3.3) of
the firing condenser before it has sufficient

lay in arming is introduced by charging the
firing condenser through a resistor. The maxi-
mum and minimum arming distances are 1,500
and 1,000 ft when a 1.5-megohm resistor and
1.0-mf firing condenser are used.

From the standpoint of safety the actual dis-
tribution of early bursts is pertinent. Field test
results give the following:

Early functions Distance to burst

per cent of total rounds fired (ft)

0 1,400

1 2,300

1.5* 3,200

* Based on 2,006 rounds.

The minimum release range (to insure arm-
ing) is a function of rocket temperature and

FUZE FOR NAVY ROCKET AR 5.0

219

plane speed. The effect of these parameters on
the minimum release range is shown in Fig-
ure 4.

Burst Heights

The average fuze is designed to function 30
to 40 ft above ground when fired from a plane
in a 40-degree dive. However, functions at 15
to 60 ft (or considerably more over water)

Figure 3. T-2004 fuze on AR 5.0 rocket.

may be expected because of variations in the
nature of the terrain and of variations among
individual fuzes.

Curves showing variation of burst height
with angle of dive are given in the data sheets
of Section 5.5. Because of the dependence of
range dispersion upon dive angle, values of 30
degrees or greater are recommended.

53,3 Structure

General Arrangement

The layout of the T-2004 is identical with
that of the ring-type bomb fuzes (see Section
5.4.3).

Arming Mechanism

The arming scheme of the T-2004 is essen-
tially the same as that of the ring-type bomb
fuzes (Section 5.4.3) except for the mechanism
that operates the slow-speed shaft and the in-
troduction of RC delay, as used in the T-6
Army rocket fuze. The slow-speed shaft is con-
trolled by a device that occupies the same space

as the gear train of the bomb fuze but which
is a combination of a gear train and an accel-
eration-operated system of levers and locks,
as shown in Figure 29, Chapter 4. To obtain
arming, the vane must turn a predetermined
number of revolutions during an acceleration
in the proper direction of more than lOp. The
spring-loaded slow-speed shaft is provided with
a toothed lever arm which is entrained with the
vane shaft by a worm gear system. If the vane
is turned too much in the absence of adequate
acceleration, the lever is forced against a lock,
and the gear train strips at such a point that
subsequent removal of the lock by acceleration
will not free the slow-speed shaft. Adequate
acceleration during the turning of the vane
removes this lock. The weight which operates
the locks is shown in the unaccelerated position
(upper left of Figure 29, Chapter 4), and the
accelerated position (upper right of figure).
As soon as the slow-speed shaft is freed from
the gear train, the spring rotates it 90 degrees.
At this position further motion of the slow-
speed shaft lever arm is blocked by a projection
on the weight. As soon as acceleration falls to
a low level, this block is removed, and the slow-
speed shaft is snapped around by the spring an

Figure 4. Minimum release range for T-2004.

additional 90 degrees into the armed position.
Since there is no further motion of the shaft,
a transfer pin is unnecessary, and the detona-
tor rotor is permanently locked to the shaft.

The fuze is mechanically armed and the fir-
ing circuit is closed at the end of acceleration
of the rocket. Detonation cannot occur until
somewhat later, however, on account of the RC
delay in the charging circuit of the firing con-
denser.

SECR

220

CATALOGUE OF FUZE TYPES

R-F System

The oscillator diode circuit of the T-2004
fuze is the same as that of the bomb fuzes con-
taining the OD circuit (see Section 5.4.3)
except for the addition of 50 mmf to the capac-
ity of all the by-pass condensers.

Amplifier

The amplifier circuit of the T-2004 fuze is
the same as that of the bomb fuzes (see Section
5.4.3), except for the addition of a 0.02-mf
condenser and 0.33-megohm resistor in series
from the pentode grid to ground, in order to
reduce the gain of the amplifier. The resistors
in the feedback network are altered to give a
suitable peak amplification frequency and a
number of minor changes in circuit component
values are introduced. These changes are shown
in Table 4 which gives values that differ from
those of the amplifier No. 11 of the T-50-E4
bomb fuze (see Table 8) .

Table 4. T-2004 amplifier component variations

from T-50-E4 amplifier 11.

Resistor

(R)

Value

(megohm)

Condenser

(C)

Value

(mmf)

0.68

20,000

0.68

500*

1,000

2,000

* See text regarding addition to circuit.

Power Supply and Firing Circuit

The power supply and firing circuit (Figure
77, Chapter 3) used in the Navy rocket fuze
is nearly the same as that used in the bomb
fuzes. The main difference is the introduction
of RC delay. The thyratron plate and one deto-
nator contact spring are connected to B-f-
through a high-resistance R27 (0.51 megohm,
nominal). The other detonator contact spring
is connected, through the firing condenser C20,
to ground. On completion of the detonator cir-
cuit, the firing condenser is charged at low
current through the detonator. The other dif-
ference, made possible by improvements in con-
denser construction, is the use of an additional
filter condenser, C23.

5 4 BOMB FUZES

5,4,1 General

Military Requirements

The military requirements for VT bomb
fuzes may be classified as (1) requirements for
GP fuzes, and (2) requirements for fuzes in-
tended to be used for specific purposes (e.g.,
the enhancement of the blast effect of a certain
bomb). Generally speaking, the requirements
of the second kind were specified after the
properties of the fuzes had been well estab-
lished, and these requirements can be described
with some accuracy. On the other hand, the re-
quirements for GP fuzes did not remain fixed
throughout the course of development, and the
description of these requirements cannot be
given completely without introducing an unde-
sirable amount of historical detail.

This is particularly true in the case of the
ring-type fuzes. For example, the initial re-
quirements of uniformity of burst heights were
such that engineering calculations indicated
that it would be necessary to manufacture
ring-type fuzes in three different carrier-fre-
quency bands in order to cover the specified
range of bomb sizes. Certain compromises were
made and the ring-type fuzes were manufac-
tured in only two carrier bands. Later, the ad-
dition of Navy requirements to the existing
Army requirements led to the manufacture of
two types of fuzes in each carrier band.

Largely as a result of knowledge gained from
effect-field tests, and service experience on the
relative usefulness of different types of bombs,
it gradually became apparent that uniformity
of burst heights for a wide range of bomb sizes
and release conditions was not so important as
had originally been supposed. In essence, it was
concluded that under many conditions the
effectiveness of an air burst was so much
greater than that of a ground burst that close
control on the height of the air burst was of
relatively minor importance. This conclusion
played an important part in the decision, made
shortly before the close of World War II, to
reduce the number of ring-type bomb fuzes in
production from four to one.

The difficulty of adequately treating this sub-

SECRET

BOMB FUZES

221

ject is further enhanced by some differences
in military opinion on the usefulness of dif-
ferent applications of the fuzes. For example,
although in certain quarters it was concluded
that the fuzes would be of little use in attacking
well-entrenched positions, the fuzes were actu-
ally used most frequently against antiaircraft
positions, and the results reported by the users
were, on the whole, very satisfactory.

For the purposes of this report, the following
general requirements are presented as adequate
to specify fuzes that are useful under a rather
wide variety of service conditions.

Reliability. The fuzes should give, on the
average, proper functions in excess of 70 per
cent when release is made at any altitude from
that required to ensure arming up to at least
20,000 ft. This performance should hold for all
train spacings in excess of a minimum deter-
mined by a reasonable area of effectiveness of
a single bomb of the train.

Burst Heights. Under the conditions just
stated, the great majority of proper functions
should lie within the range 10 to 100 ft above
the target area. This implies that the average
height of proper functions should lie within
the range 15 to 50 ft above the target area,
regardless of release conditions within the
limits stated above.

Safety and Arming. The fuzes should be at
least as safe to use as the safest of all other
types of bomb fuzes. None should arm before
traveling the minimum safe air travel [Min-
SAT] specified for the particular type of fuze,
and practically all should arm within ±10 per
cent of the mean air travel to arming of the
particular bomb-fuze combination under con-
sideration.

General Types

As mentioned above (see Section 5.1.2)
bomb fuzes may be divided into two major
groups on the basis of the antenna system — the
ring type and the bar type. Another type of
classification is based on the r-f circuit. In this
system, the classification depends upon whether
detection is accomplished by a tuned diode de-
tector [OD], by a reaction grid detector [RGD]
or by a power-oscillating detector [POD] . This
section will explain both types of classifica-

tion and describe the general operational
characteristics of each group.

Ring and Bar. The external differences be-
tween the ring and bar type are shown in Fig-
ure 5, Chapter 1. The antenna system of the
ring type consists of the ring of the fuze to-
gether with the body of the bomb. This type of
excitation is known as longitudinal excitation.
The antenna of the bar type consists of the two
bars on the fuze, and does not theoretically in-
clude the bomb itself. Actually, there is usually
some slight excitation of the bomb. This type
of excitation is known as transverse. See table
below for listing of fuzes of the two types.

Bar-type fuzes may be expected to give better
scores (less random functions) than most ring-
type, for two reasons. (1) Since the vehicle is
relatively unexcited in the bar type, any me-
chanical disturbance, such as fin flutter, will
affect the radiation only very slightly; and (2)
bar-type fuzes have either RGD or POD cir-
cuits, which, as will be shown later, are less
susceptible to noise disturbances than the OD.

One of the most marked differences between
the two types is the much greater burst height
possible with the bar type. This difference is
due to the orientation of the radiation direc-
tivity pattern, which differs by 90 degrees for
the two types. For an antenna short compared
with a wavelength, the maximum radiation is
at an angle of approximately 90 degrees with
the axis of the antenna, and the minimum along
the axis. Since the antenna of the bar type is
perpendicular to the axis of the bomb, the
maximum radiation is along the bomb axis.
Therefore, the bar-type fuzes have their maxi-
mum sensitivity directly forward. Since for
most release conditions the angle the bomb
makes with the vertical is very small, this high
forward sensitivity aids in the obtaining of
large function heights. Ring fuzes, on the other
hand, have their maximum radiation roughly
perpendicular to the bomb axis and low radia-
tion at the small angles encountered in level
flight release from high altitudes. The ring
fuzes are therefore sensitive to objects they
pass while the bar fuzes are sensitive to objects
directly ahead. As a result, bar fuzes give much
higher heights for level-flight release conditions
when the bomb is close to vertical.

sec:

222

CATALOGUE OF FUZE TYPES

Key to bomb fuzes

Ring type

Bar type

Brown

White

Yellow

White

OD circuit

RGD

POD

T-50-E1

T-50-E4

circuit

T-89

T-90

T-51

T-82

T-91

T-92

T-51-E1 (M-166)

Series

RGD circuit

T-51-E2

T-91-E1 (M-168) T-92-E1

T-712

R-F Circuit . The operational differences be-
tween fuzes having the OD, RGD, or POD cir-
cuits are less marked than those based on the
antenna system, but they are significant. In gen-
eral, better scores may be expected from RGD
and POD than from OD circuits. The tuned
diode detector circuit is much more sensitive to
frequency modulation (usually produced by
microphonics) than the RGD or POD circuits.
Another factor influencing OD performance is
tuning of the OD circuit, which is different on
each bomb. Hence the diode circuit cannot be
perfectly tuned on every vehicle. Since both r-f
sensitivity and noise-response depend on tun-
ing, there will be more spread in both height
and scores for OD units on different vehicles
than for RGD.

The RGD circuit, therefore, is less suscep-
tible to noise disturbances, particularly tube
microphonics, than the OD and performs more
uniformly on different vehicles.

Specific Applications

VT bomb fuzes may be divided into two cate-
gories, depending upon mode of flight during

release. Three fuzes, T-91, T-91-E1, and T-92
were designed with short arming time (2,000
to 2,600 ft MinSAT) specifically for dive bomb-
ing and low-release altitude; all other fuzes
(with 3,600 ft or greater MinSAT) for level
flight release.

Specific applications of these fuzes may be
divided into two groups: (1) those depending
upon the effect of the blast associated with the
bomb burst, and (2) those depending upon the
effectiveness of fragments from the bomb and
its contents. In Table 5 (a) and (b) are of the
first type and (c) through (f) the latter. The
recommended type of fuze and most desirable
burst height are given along with estimates of
effectiveness as compared with similar use of
a contact fuze.

With the exception of the first and last appli-
cations listed below, the M-166 and M-168
fuzes may be regarded as adequate to meet all
the applications listed. The M-166 is suitable
when large burst heights are needed and the
M-168 for low burst heights or for dive-bomb-
ing applications. The earlier models have been
listed merely for completeness.

5 4.2 Functioning Characteristics
Safety and Arming

MinSAT s Available , Normal Arming. VT
bomb fuzes have been designed with minimum
safe air travels of 2,000, 2,600, 3,100, 3,600,
4,500, and 8,000 ft. Of these MinSATs, how-

Table 5. Applications of bomb fuzes.

Application

Burst

height

(ft)

Estimated

advantage

Fuze

Blast effect, M-56

40-70

1.5 to 2

T-712

Mine clearance by blast from Mk-44

10-20

Up to 2

T-50 type

General purpose (antipersonnel and light materiel)

1. For 500- and 1,000-lb bombs

20-70

1 to 20

T-90, T-50-E4, T-92*

2. For bombs smaller than 500 lb and others up to
2,000 lb

20-70

1 to 20

M-166, M-168, T-50-E1,

3. For all vehicles

20-70

1 to 20

T-89, T-91

M-166, T-51-E2

Chemical warfare (gas)

100-200

4 to 7

M-166, T-51-E2

Fire bombing

1. For 165-gal belly tank

40-80

M-166, T-51-E2

2. For M-10 spray tank

5-15

T-90, T-50-E4

* The T-92 is listed last on account of inferior reliability (see Section 5.4.2).

SECRET

BOMB FUZES

223

ever, only three were used on fuze types reach-
ing the production stage. These were :

MinSAT Production type

2,000 ft T-91, T-91-E1 (M-168) ; T-82-E2

2.600 ft T-92

3.600 ft T-50-E1, T-50-E4 ; T-51-E1 (M-166), T-51-

E2; T-82-E1 ; T-89, T-90

Data on the safe vertical drop and minimum
release altitude corresponding to different
MinSATs under various conditions are tabu-
lated in reference 7. The types with 2,000-ft
and 2,600-ft MinSATs were intended spe-
cifically for dive-bombing applications.

MinSATs Available, Delayed Arming. The
air travel to arming on all the above types
(except the T-82-E1) can be extended by the
use of the device arming delay, air-travel, M-l
(formerly T-2-E1), which will provide up to
about 20,000 ft of additional air travel to arm-
ing for any fuze adapted for holding it. The
arming delay device is a wind-driven vane
mechanism which is clamped on the fuze in
such a way as to prevent rotation of the fuze
arming vane (see Figure 1, Chapter 4). When
the delay vane has rotated through a preset
number of revolutions (manually adjustable in
the field), the delay disengages itself from the
fuze, allowing the fuze arming mechanism to
operate in the normal fashion. The arming
delay is constructed with a setting dial con-
taining 25 divisions, each of which corresponds
to about 800 ft of air travel on the M-30 test
bomb. Data on arming delay settings under var-
ious conditions are given in reference 7.

Effect of Bomb Size on Air Travel. For a
given rotor setting the air travel to arming
generally increases with the bomb size. This
effect is due to the additional obstruction
offered by the larger bomb nose to the flow of
air past the vanes. The relative air travel for
different vane types on various vehicles is as
follows :

Vane

Bomb

Metal

M-30

M-81

M-64

and

100-lb GP

260-lb frag

500-lb GP

plastic

1.00

1.02

1.15

Vane

Bomb

Metal

M-65

M-66

M-56

and

1,000-lb GP

2,000-lb GP

4,000-lb GP

plastic

1.32

1.58

1.48

Safety Pin. Booster cups on the production-

type fuze models, with the exception of
T-50-E1, T-50-E4, and T-51-E2, are equipped
with a safety pin for indicating that the fuzes
are safe for handling. The pin locks the arming
mechanism and can be inserted only if the det-
onator rotor is in the unarmed position. Fuzes
are issued with the safety pin in place and can-
not be installed in the fuze-well unless the pin
is removed.

Reliability

With few exceptions, ring-type fuzes released
under standard test conditions (from 10,000 ft
at 200 mph), can be expected to give uniform
performance of about 75 to 80 per cent proper
functions. The M-168 fuzes should give con-
sistently better scores (about 90 per cent
proper). The performance of T-92 fuzes was
not satisfactory (an average of only 58 per
cent proper functions was obtained in experi-
mental and acceptance tests). A possible ex-
planation for the increased incidence of early
functions may be found in the fact that the
T-92’s were built with broader pass-band am-
plifiers and were more sensitive than the other
fuzes. The fuze appeared to be unduly suscep-
tible to certain structural variations in stand-
ard bombs. The effect of fin thickness on fuze
performance is discussed in Chapter 9.

Fuze performance is dependent to some ex-
tent upon altitude of release. Scores will gen-
erally be slightly poorer for release altitudes
above 10,000 ft than under standard release
conditions; on the other hand, performance
can be expected to improve somewhat with
lower release altitudes.

Bar-type fuzes, on the whole, should give
better performance than ring-type fuzes, ex-
cept the M-168. Under standard release condi-
tions, proper function scores of T-51 and T-82
type fuzes should fall close to an average of 90
per cent.

The possibility of sympathetic functioning
must be considered when proximity fuzes with-
out arming delays are released in train. If the
spacing between bombs is too small, one armed
fuze may react upon the random burst of an-
other. The train spacing should exceed 50 ft
on the ground for small bombs (less than 500
lb), and should exceed 100 ft for large bombs.

224

CATALOGUE OF FUZE TYPES

Train spacing is unimportant if arming delays
are used because the delays postpone arming
until the bombs are widely spaced.8 These de-
vices can be used on most models (see Section
5.4.2).

Burst Heights

Burst heights of VT bomb fuzes are affected
by a number of factors external to the fuze,
such as vehicle, altitude of release, plane speed,
and target factor.

Ring Type. Proper function heights of ring-
type fuzes under standard test conditions (i.e.,
dropped over water from 10,000 ft at 200 mph)
can be expected to lie between 10 to 100 ft, with
average heights ranging from 15 to 50 ft. Satis-
factory uniformity of burst heights in the same
range should be obtained for fuze-vehicle com-
binations recommended in the fuze data sheets
below. Lower function heights will result from
the mismatching of vehicle and fuze. The effect
of release altitude upon burst height is not
simply defined. For some fuzes, such as the
T-89, averages of proper function heights can
be expected to increase with increasing alti-
tudes of release; for other fuzes, the reverse is
true (see Figures 9 and 21).

Bar Type. The forward sensitivity pattern
of bar-type fuzes causes much higher burst
heights than those obtained with the ring-type
fuze. Function heights as high as 240 ft above
water are considered proper for T-51 and T-82
type fuzes, released under standard conditions.
Average function heights of 50 to 100 ft over
land can be expected from either fuze mounted
on M-81. Theoretically, performance of bar-
type fuzes should be nearly independent of
vehicle; however, lower burst heights will be
produced by fuzes mounted on 500-lb bombs
and larger (with the exception of the T-51 on
M-56) than by fuzes mounted on 100- to 260-lb
bombs. For burst heights of bar-type fuzes on
various bombs relative to burst heights on
M-81, see Table 14.

343 Structure

General Arrangement

The general layout of all bomb fuzes (except
the T-82 series) is illustrated in Figure 17,

Chapter 4. Except for the much larger an-
tennas (ring or dipole bars) and the introduc-
tion of an axially located drive shaft running
from the vane through the electronic system,
the front section is practically identical with
that of the Army rocket fuzes T-5 and T-6. The
battery of the T-5 is replaced by a magneto-
type generator, a gear-reduction system, and
most of the components of a rectifier-filter sys-
tem for the generator, all in about two-thirds
of the fuze length required for a battery. At
this level, the fuze housing is shouldered in to
seat on the nose of the bomb. The narrower
extension of the housing contains a cylindrical
filter and firing condenser, through the center
of which runs a slow-speed shaft from the gear-
reduction system to the terminal elements of
the arming mechanism. The housing is closed
at the rear end by a threaded cup containing
a booster charge of tetryl.

All bomb fuzes are shipped completely
assembled and loaded and require no field test-
ing or assembly operations such as those re-
quired with the battery-powered T-5 and T-6.

Arming Mechanism

The principal elements of the arming mech-
anism of all bomb fuzes (except the T-82
series) are shown in Figure 20, Chapter 4. A
worm-gear train drives a slow-speed shaft to
which is keyed a Bakelite detonator rotor by
means of a spring-loaded transfer pin. The
detonator rotor carries an electric detonator,
eccentrically located in a hole as shown in the
rear-end view, Figure 24, Chapter 4. The rotor
rides in a Bakelite housing which is mounted
on the rear end of the rectifier filter assembly
shown in Figure 27, Chapter 4. This housing
serves two purposes : it has a slot that permits
the spring-loaded transfer pin to leave the slot
in the slow-speed shaft after the rotor has
turned a predetermined angle, thus locking the
rotor in the armed position; it carries electric
contact springs that complete the detonator
circuit immediately before the rotor is locked in
place. In the unarmed position, the tetryl
booster charge is protected from the detonator
by a thick brass safety plate. This plate carries
a tetryl plug to which the detonator is juxta-
posed in the armed position.

BOMB FUZES

225

All bomb fuzes are provided with a vane-
locking pin, or equivalent device, from which
is withdrawn an arming wire when the bomb is
dropped (see Figure 20, Chapter 4).

A safeguard that was necessary on the
earlier models was the closed slot on the slow-
speed shaft. In order to prevent assembly with
an incorrect rotor setting, the keyway is not
extended to the end of the slow-speed shaft.
The rotor housing is notched at the proper
angle from the armed position, and the transfer
pin can be pressed into the keyway of the slow-
speed shaft only when this keyway is aligned
with the notch. An incorrect setting can be
obtained in these models only by rotation of the
vane, the locking pin of which was sealed in
place before assembly.

This safeguard was unnecessary in the later
models, which are provided with a safety pin
(see Figures 23 and 24, Chapter 4) which
can be inserted into and withdrawn at will
from the completely assembled fuze only if the
detonator rotor is in the correct position.

The data sheets (see Section 5.5) tell which
of these additional safety features appear in
each of the production model fuzes. The possi-
bility of using the arming delay device is also
indicated in Section 5.5. The purpose and
method of using this auxiliary device has
already been outlined in Section 5.4.2 above.
Its construction and operation can be readily
visualized from Figure 1, Chapter 4.

R-F System

Oscillator-Diode [ OD ] Circuits. The OD cir-
cuit used in bomb fuzes is nearly the same as
that used in the Army rocket fuzes T-5 and T-6
(see Figure 2). In the bomb-fuze diode circuit,
R5 is connected to the other side of C4 instead
of to ground. No switches are used in the A and
B supply lines, the resistance in the diode fila-
ment line is increased to 10 ohms, and the ca-
pacity of the by-pass condensers C2, C5, and
C6, is increased to 150 ppf each. Another 150-
|i|if by-pass from the B supply line to ground
is located in the amplifier section of the fuze.

The Reaction-Grid-Detector [ RGD ] Circuit.
The RGD circuit is used in the M-168 and the
T-92-E1 ring-type bomb fuzes (see Figure 5).
Component values are given in Table 6.

Table 6. Component values for RGD oscillator
in M-168 and T-92-E1 fuzes.*

Value

Resistor

(ohms)

Condenser

(w*f)

100,000

47,000

2,200

150

C22

150

Triode :

: NR3A

* Ll and L2

are adjusted to obtain required frequency and sen-

sitivity. C22 is

located in the amplifier section.

The RGD oscillator used in the M-166 bar-
type bomb fuze is shown in Figure 6. Values of

components are given in Table 7.

Table 7.

Component values of RGD

oscillator

in M-166 fuze.*

Value

Resistor

(ohms)

Condenser

(wrf)

1,000

33,000

47,000

Triode

: NR3A

* Ll : oscillator and antenna coils wound on same core. L2 : r-f
choke.

Figure 5. Reaction-grid-detector circuit used in
M-168 and T-92-E1 bomb fuzes.

Oscillator Assemblies. With the exception of
the T-82, the physical layout of the oscillators
in all bomb fuzes is nearly the same. The
mounted components and their circuital rela-
tionships are illustrated in Figures 13, 14, and
15, Chapter 3. Phenolic (thermosetting)
mounting blocks (Figure 5, Chapter 6) were
used in all except the M-166, in which a sty-

SECRET

226

CATALOGUE OF FUZE TYPES

ramie (thermoplastic) block was used (Figure
6, Chapter 6).

Amplifier

Circuits in Ring-Type Fuzes. The basic am-
plifier circuit of all ring-type bomb fuzes is
shown in Figure 26, Chapter 3. Table 8 pro-

Table 8. Component values for amplifier No. 11
in T-50-E4 fuze.*

Resistor

Value

(megohms)

Condenser

Value

<w*f)

0.68

5,000

5.6

3-20

RIO

1.0

CIO

5,000

Rll

1.5

Cll

500

R12

0.47

C12

200

R13

2.2

C13

200

R14

6.8

C14

1,000

R15

2.2

C15

500

R19

3 ohms

C16

R21

6.8 ohms

R30

1.0

Pentode :

NS-5

Thyratron

: NS-4

* C16 is optional; may be connected between ground or filament
and any part of the input circuit to minimize r-f effects. Additional
items in amplifier section: a test lead brought out from the thyratron
grid; 150-.u/u£ by-pass condenser from B+ to ground.

vides the component values for the No. 11 am-
plifier of the T-50-E4 and notes on minor dif-
ferences from the circuit shown in the figure.
Table 9 shows the differences that occur in
other amplifiers of this series.

Figure 6. Reaction-grid-detector circuit used in
M-166 bomb fuze.

Circuit in Bar-Type Fuze. The basic amplifier
circuit of the M-166 (T-51-E1) bar-type bomb

fuze is shown in Figure 30, Chapter 3, for
which Table 10 provides the component values.

Table

9. Variations from

amplifier

No. 11

(T-50-E4) in other ring- type bomb fuzes.

Fuze

T-50-E1

T-91

T-89

M-168

T-92

T-92-E1

Amplifier

No. 10

No. 20

No. 16

No. 18

Resistor

(values

in meg-
ohms)
Rll

1.68*

R12

Condenser

0.82

1.5*

0.82

0.68*

(values
in /ifi f )

10,000

2,000

5-20

0-20

CIO

20,000

Cll

200

C12

500

500$

C13

500

C14

2,000

C15

300$

200§

* Adjust to obtain required frequency.

f Feedback loop connection is shifted from pentode plate to the
thyratron grid side of C14.

t A 200-fifif condenser is connected from the input line to ground.
§ Feedback loop resistors Rll and R12 connected to center point
between the legs of pentode filament by two 1,000-ohm resistors.

Table 10. Component values for amplifier in
M-166 (T-51-E1) fuze.*

Resistor

Value

(megohms) Condenser

Value

<W*f)

3 ohms

0.39

0.005

1.0

50 nni

0.47

0.01

1.0

0.0002

5.6

0.0004

R10

2.2

0-20 w f

Rll

3 ohms

Pentode: NS-5, NR-5 or

NGE-5

Thyratron :

NS-4

* Not shown

in Figure 30,

Chapter 3: A

test lead brought out

from the thyratron grid; each side of the A supply line is grounded
(in the oscillator section) by a 3-ohm resistor shunted by a 50-^f
by-pass condenser.

Amplifier Assemblies. A number of different
types of amplifier assemblies are found in the
bomb fuzes. These are illustrated in the follow-
ing figures in Chapter 6 : Figure 17, right, for
Philco models; Figure 17, left, for Emerson
models (both are disk variations of the sand-
wich type of assembly) ; Figure 14 for the
Zenith M-166 (“dog collar” type of assembly).

PRODUCTION FUZE DATA SHEETS

227

For reasons discussed in Chapter 6, the type of
assembly used is, generally speaking, a charac-
teristic of the manufacturer rather than of the
fuze, and a more detailed description is unwar-
ranted here.

In all bomb fuzes, the thyratron is included
in the amplifier assembly. Tung oil was used as
a potting material for the amplifier assemblies,
except for late Emerson production, for which
Glidden potting compound was used.

Power Supply and Firing Circuit

The power supply and firing circuit used in
bomb fuzes is shown in Figures 75 and 76,
Chapter 3. The lead from the thyratron plate
[TP] tap is connected to one of the spring con-
tacts in the detonator rotor housing; the other
spring contact is connected to B+ and through
the firing condenser C20 to ground. The filter
condenser is C19, and C18 is the voltage regula-
tion condenser.

55 PRODUCTION FUZE DATA SHEETS
5,51 Scope

The following set of data sheets covers infor-
mation, where available or pertinent, for pro-
duction fuzes only, in the following order :

Item 1. Tabulations of arming, electrical,
and performance data.

Item 2. Curves of burst-height performance.

Item 3. Amplifier gain curves.

Item 4. Radiation patterns and loading
curves.

The T-51-E2 has been omitted, since it repre-
sents only about 4 per cent of the total bar-type
production and its characteristics were prac-
tically the same as those of the Zenith M-166,
except that it lacked the safety pin.

The T-712, although produced on an even
smaller scale than the T-51-E2, has been in-
cluded because certain of its characteristics are
quite different from those of other bar-type
fuzes. It is understood that its production was
limited on account of the limited supply of M-56
bombs.

Coverage in this section has been limited to
production items because these are the only

fuzes which are in stock in appreciable quanti-
ties.

Explanatory Notes

Item 1.

(a) In the tabular data , entries are omitted
if all in a row are repetitions of that in the
first column.

(b) Electrical data represents overall produc-
tion averages as shown on CTL Quality Control
Charts, where available.

For longitudinally excited fuzes, the maxi-
mum sensitivity is given ; the approximate load
resistance R (in 103 ohms) at which maximum
sensitivity occurs appears in parentheses after
the sensitivity value. For OD circuit fuzes, the
sensitivity S' at any other load R' can be calcu-
lated with sufficient accuracy from the formula
S' 4 RRf

S (R + R'f

For the RGD circuit, this formula is less accu-
rate, and loading curves are given where avail-
able.

The laboratory data on detuning apply to OD
circuit fuzes as tested with a standard load
approximately equivalent to that presented by
the missile for which the fuze was designed.
For bomb fuzes, the bombs represented in the
laboratory tests were M-30 for Brown fre-
quency and M-64 for White. For information
concerning the effect of detuning on sensitivity,
see Section 2.7.

Effective critical voltages are the maxima
obtained in the detuning tests.

(c) Function scores are averaged without
regard to reflection coefficient or plane speed
but are restricted to missiles, as stated in
Table 11. The very few late functions are in-
cluded with the proper functions.

Function heights are for release from 10,000
ft at a plane speed of 200 mph over water with
a reflection coefficient of approximately 0.81,
unless otherwise indicated by footnote. Part of
the original data were obtained under other
conditions. The method of reduction to a com-
mon condition is covered in Section 9.4.

(d) Under production , the quantities given
are the approximate number of metal parts
(MP) lots produced, usually about 1,000 units
per lot.

228

CATALOGUE OF FUZE TYPES

(e) The AN/CPQ-( ) designation system
was originally set up to distinguish between
vane leads, arming distances, rotor settings,
etc., in different metal parts assemblies as de-
livered from the factory. Later it became neces-
sary to make changes in the rotor settings in
assembling some of the metal parts products
into complete fuzes at Picatinny Arsenal, so
that the AN/CPQ designations lost their sig-
nificance in some cases. However, since this
nomenclature was used almost exclusively in
the many reports of the Control Testing Lab-
oratory at the National Bureau of Standards,
it has been recorded here as an aid to anyone
who has occasion to study the laboratory per-
formance of production models. The lack of a
complete 1-to-l correspondence between metal
parts designations and fuze (T- or M-) desig-
nations is unavoidable.

Item 3. The PkAF and peak gain as appear-
ing on the curves are not always consistent
with the audio-frequency at peak amplification
[PkAF] and millivolts to fire [MvF] at peak
given in the tabular data, because considerably
smaller samples were used in determining the

amplifier characteristics. The curves can be
relied upon for shape but should be adjusted
for location of peak.

Item U- Only those radiation patterns that
are most likely to be useful in the calculation of
burst heights have been included; for other
patterns and much useful additional data for
this purpose, see reference 3.

Abbreviations.

MinSAT: Minimum safe air travel, dur-
ing which no fuzes will arm.

CF: Carrier frequency of the fuze trans-
mitter.

S: Sensitivity, as defined in Section 3.1.2.
Amp. No. : Signal Corps identification
number of amplifier.

PkAF: Audio-frequency at peak amplifi-
cation.

MvF: Millivolts to fire (the thyratron)
applied at amplifier input.

CV : C-bias voltage on the thyratron.
EC: Effective critical voltage, at which
bias voltage the thyratron fires.

Rel gain : Relative gain of the amplifier.
SD time: Self-destruction time.

Data Common to Production VT Bomb Fuzes

Table 11. General purpose models.

Fuzes

Ring-type Bar-type

Brown

White

Property

T-50-E1, T-89, T-91,
T-91-E1 (M-168)

T-50-E4, T-90, T-92

T-51-E1 (M-166)

Physical characteristics:

Overall length (in.)

10t?

10^2

Length from shoulder (in.)

Overall width (in.)

Weight (lb)

Vane speed range* (1,000 rpm)

15-30

20-35

Proof test conditions:

Bomb

M-64

M-57, -81, -88

Release alt. (ft)

10,000

Uses:

Physically interchangeable with contact fuze

M-103

Some bombs for which the fuzes are useful

GP: M-30, -57, -66
Frag: M-81, -88

GP: M-64, -65

GP: M-30, -57, -64,
-65, -66

Frag: M-81, -88

* Laboratory test speed range.

PRODUCTION FUZE DATA SHEETS

229

5.5.2

Bomb Fuzes, Ring Type, Brown Carrier

Table 12. Characteristics and scores.

M-168

(T-91-E1)

Level or dive release

T-91

Level release

T-89 T-50-E1

Arming

MinSAT (ft)

2,000

3,600

Safety pin

Yes

Delay device

Yes

Yes*

Rotor setting (°)

110

100

Rotor shaft

Open

Closed

Vane

10-blade metal prop

Vane angle (°)

Electrical

Radio

Circuit

RGD

+0.9

-1.5

+ 1.1

-1.5

— 1.5f

S (volt)

30 (5)

16 (5.5)

18 (5.5)

16.5 (5.5)

Detuned (%)

Audio

Amp. No.

PkAF (c)

116

MvF (Pk)

CV (-volt)

7.7

8.2

7.6

8.4

3.8

4.9

4.3

4.8

Proof performance

Burst height (ft)

Proper (%)

Random (%)

Dud (%)

4 1

Production

Manufacturer

Emerson

Philco

Quantity (MP lots)
AN/CPQ-

130

PA-

329

307

263

180

* Not loaded with fuzes, f First 50 MP lots manufactured at + 2, excluded from average. + 19% dud on first sample tests of the
first 42 lots, due to faulty rotor contact spring adjustment. Not included in average.

LOAD RESISTANCE, RA(I030HMS)

Figure 7. Oscillator loading characteristics of
M-168 bomb fuze. Plate current IVy grid voltage
Eg , and radiation sensitivity 5 are shown as
functions of radiation resistance Ra.

Figure 8. Amplifier gain as function of signal
frequency for ring-type Brown-carrier bomb fuzes.

SECRET

RELEASE ALTITUDE (10 FT) M ^ _ RELEASE ALTITUDE (10 FT)

230

CATALOGUE OF FUZE TYPES

200 250 300 350 400

PLANE SPEED (M PH)

?IGUR1
?or M
'eflect

E 9. Iso-burst-height curves (predicted)
-168 fuze on M-64 (500-lb) GP bomb for
ion coefficient of 0.5.

200 250 300 350 400

PLANE SPEED (M PH)

Figure 11. Iso-burst-height curves (predicted)
for M-168 fuze on M-81 (260-lb) fragmentation
bomb for reflection coefficient of 0.5.

Figure 10. Iso-burst-height curves (predicted)
for T-50-E1 or T-89 fuze on M-81 (260-lb) frag-
mentation bomb for reflection coefficient of 0.5.

Figure 12. Cumulative distribution of indi-
vidual burst heights for various ring-type Brown-
carrier fuzes on M-81 (260-lb) fragmentation
bomb. Reflection coefficient is 0.8 except as noted.
A, T-50-E1 and T-89; B, Philco T-91, reflection
coefficient 0.6; C, GE T-91; D, M-168, reflection
coefficient 0.6, plane speed 240 mph.

PRODUCTION FUZE DATA SHEETS

231

Figure 13. Radiation directivity pattern for
Brown frequency longitudinal end excitation of
bombs: A, M-30 or M-81 (V2 wp = 0.124); B,

M -66 (% irp = 0.207).

0 (DEGREES)

Figure 14. Small angle detail for M-30 pattern
of Figure 13 for frequencies: A, B — 1; B,
B + 5.5.

2 4 6 8 10 12 14 16 18 20

0( DEGREES)

Figure 15. Small-angle detail for M-81 pattern
of Figure 13 for frequencies: A, B + 0; B,
B + 9.4.

0 2 4 6 8 10 12 14 16 16

0 (DEGREES)

Figure 16. Small-angle detail for M-66 pattern
of Figure 13 for frequencies: A, B + 5.5; B,
B — 0.8.

CRE

232

CATALOGUE OF FUZE TYPES

Bomb Fuzes, Ring Type, White Carrier

Table 13. Characteristics and scores.

Level release Dive release

T-90 T-50-E4 T-92 T-92-E1

Arming

MinSAT (ft)

Safety pin
Delay device
Rotor setting0
Rotor shaft
Vane

Vane lead (in.)
Electrical
Radio
Circuit
CF

S (volt)

Detuned (%)
Audio
Amp. No.

PkAF (c)

MvF (Pk)

CV (—volt)

EC (-volt)

Proof performance
Burst height (ft)
Proper (%)
Random (%)

Dud (%)
Production
Manufacturer
Quantity (MP lots)
AN/CPQ-
PA-

3,600

Yes

Yes*

145

140

Closed

3-blade Bakelite prop

. OD

+9.0

+9.3

19 (6.5)

18 (6.5)

190

185

7.4

4.6

Emerson

264

181

2,600

Yes

110

Closed

Open

RGD

+8.5

+5.6

18 (6.5)

30 (3.5)

160

156

7.4

7.6

4.6

4.0

1A, IB

306

* Not loaded with fuzes. Requires mounting bracket as on T-51.

Figure 17. Cumulative distribution of individual
burst heights for ring-type White-carrier fuzes
on M-64 500-lb GP bomb. Reflection coefficient
is 0.8.

Figure 18. Amplifier gain as function of signal
frequency for ring-type White-carrier bomb
fuzes. Amp 11 in T-50-E4, Amp 16 in T-92, Amp
18 in T-92-E1.

(eu-

PRODUCTION FUZE DATA SHEETS

233

Figure 19. Radiation directivity patterns for
White + 10 frequency longitudinal end excitation
of bombs : A, M-64 ( V2 tt/3 = 0.208) ; B, M-65
(y2 tt/3 = 0.182).

O 2 4 6 8 10 12 14 16 18 20

6 (0EGREES)

Figure 20. Small-angle detail for Figure 19:
A, M-65 at W + 10; B, M-65 at W + 0.2; C,
M-64 at W + 10; D, M-64 at W + 0.2.

200 250 300 350 400

PLANE SPEED (M P H)

Figure 21. Iso-burst-height curves (predicted)
for T-50-E4 or T-90 fuze on M-64 (500-lb) GP
bomb for reflection coefficient of 0.5.

SECRET

234

CATALOGUE OF FUZE TYPES

Figure 22. Oscillator loading characteristics of the T-92-E1 bomb fuze. Plate current IP, grid voltage
Eg, and radiation sensitivity S are shown as functions of radiation resistance Ra.

Bomb Fuzes, Bar Type, Yellow Carrier

Table 14. Characteristics and scores.

Level or dive release Special for M-56 GP

M-166 (T-51-E1) T-712

Arming

MinSAT (ft)

Safety pin
Delay device
Rotor setting (°)
Rotor shaft
Vane

Vane lead (in.)
Electrical
Radio
Circuit
CF

S (volt)

Audio
Amp. No.

MvF (165 c)

MvF (300 c)

EC (-volt)

CV (-volt)

Proof performance
Burst height
Proper (%)
Random (%)

Dud (%)
Production
Manufacturer
Quantity (MP lots)
AN/CPQ-
PA-

3,600

Yes

153

Closed

3-blade Bakelite prop
6

RGD

8.5

7.4

9.9

3.7

3.8

3.6

7.9

7.8

110

50*

100

Zenith

Emerson

Zenith

230

283

♦Tested on M-81 (reflection coefficient, 0.6; speed: 240 mph).

PRODUCTION FUZE DATA SHEETS

235

Bomb weights and relative burst heights for M-166.

Bomb

Weight

(lb)

Relative

burst

height

Bomb

Weight

(lb)

Relative

burst

height

M-30

100

1.28

M-64

500

0.73

M-88

220

1.00

M-65

1,000

0.69

M-81

260

1.00

M-66

2,000

0.40

M-57

250

1.00

M-56

4,000

1.37

Figure 25. Amplifier gain as function of signal
frequency for bar-type Yellow-carrier bomb
fuzes.

Figure 23. Burst height as function of altitude
of release for Zenith T-51-E1 fuze on several
bombs. Reflection coefficient is 0.8.

Navy Rocket Fuze, Ring Type,
Brown Carrier

Figure 24. Cumulative distribution of individual
burst heights for Zenith T-51-E1 fuze on M-81
(260-lb) fragmentation bomb. Reflection coeffi-
cient is 0.6.

Table 15. Characteristics and scores.

Plane to Surface
T-2004

Arming

MinSAT (ft)

1,000

Safety pin

Yes

Vane

10-blade metal prop

Vane angle (°)

Electrical

Radio

Circuit

+ 1.3

S (volt)

Detuned (%)

Audio

PkAF (c)

125

MvF (Pk)

109

CV (-volt)

7.7

EC (-volt)

4.1

Proof performance

Burst height*

Reflection coefficient

0.81

Proper (%)

Random (%)

Dud (%)

Production

Manufacturer

Philco

Quantity (MP lots)

MP designation

AN/CPQ-3A

PA-

315

* Fired from a ground launcher at approximately 30-degree eleva-
tion. External physical dimensions are same as those of ring-type
bomb fuzes.

SECRET

236

CATALOGUE OF FUZE TYPES

HORIZONTAL FLIGHT- RELEASE ALTITUDE (FT)

1
CL

LlI

300

00 UJ

♦ H

2?°5

30°DIVE ANGLE

HORIZONTAL FLIGHT-RELEASE ALTITUDE (FT) x

45° DIVE ANGLE

HORIZONTAL FLIGHT-RELEASE ALTITUDE (FT)

300 £

250

Figure 26. Equivalent altitude and plane speeds for level-flight and dive-bombing releases at dive
angles of 30°, 45°, and 60°. Given burst height as function of level-flight release altitude and plane speed
(see, for example, iso-burst-height curves), this chart may be used to determine burst height for dive-
bombing releases. Scale for M-64 bomb may be used for larger bombs in GP series. Scale for M-81 bomb
gives rough approximation for M-30.

PRODUCTION FUZE DATA SHEETS

237

O 10 20 30 40 50 60 70

DIVE ANGLE

Figure 27. Burst height as function of dive
angle for T-2004 fuze on 5.0-in. AR Navy rocket
for firing at plane speed of 300 knots at range of
1,500 to 2,000 yd. (Fired over ground at
Inyokern.)

Figure 28. Cumulative distribution of indi-
vidual burst heights for T-2004 fuze on 5.0-in.
AR Navy rocket for firing at plane speed of 300
knots at range of 1,500 to 2,000 yd (Inyokern
data).

Figure 29. Amplifier gain as function of signal
frequency for the T-2004 ring-type Brown-car-
rier Navy rocket fuze.

Figure 30. Radiation directivity pattern for
Brown frequency longitudinal end excitation of
5.0-in. AR Navy rocket (Vz wp = 0.130).

238

CATALOGUE OF FUZE TYPES

Army Rocket Fuzes

Table 16. Characteristics and scores.

Type: longitudinally excited. Carrier: Red, Yellow, Green.
T-5: Plane- to-plane or plane- to-ground
T-6: Ground-to-ground

Arming: MinSAT (ft) T-5: 525

T-6: 2,400

Electrical (same for T-5 and T-6)

(volt): 18.

Test Voltages

A: 1.40

A: 1.15

B: 135

B: 115

PkAF (c)

121

MvF (Pk)

Rel gain at 20 c (%)

Rel gain at 300 c (%)

SD time (sec)

Proof performance

T-6

T-5

Philco

Friez

Proper (%)

Early (%)

Late or Mid (%)

Dud (%)

Weight and dimensions (same for T-5 and T-6)

Length (overall): 7^ in. Length (outside rocket): 2^ in.
Width (overall): 3^ in. Weight (lb): 2f
Physically interchangeable with contact fuze PD-M-4

West-

Manufacturers Emer- ing-

son Friez GE Philco house All

Quantity (MP lots) 100 25 80 1 10 80 395

T-5 ON ROCKET T-22

(EGLIN FIELD ST 2-45-16)

Figure 32. Burst height as function of dive
angle for T-5 fuze on T-22 Army rocket for fir-
ing at plane speed of approximately 300 mph at
range 700 to 1,000 yd. (Fired over ground at
Eglin Field.)

Circuit: OD. Sensitivity

Figure 31. Cumulative distribution of indi-
vidual burst heights for T-6 fuze on 4.5-in. Army
rocket for several firing elevations as indicated.
Reflection coefficient is 0.96.

Figure 33. Radiation directivity pattern for
Red and Green frequency longitudinal end ex-
citation of 4.5-in. Army rocket.

SECRET

PREPRODUCTION FUZES

239

Figure 34. Amplifier gain as function of signal
frequency for the T-5 or T-6 Army rocket fuze.

PREPRODUCTION FUZES

The VT fuzes which were not in production
at the end of World War II are covered in this
section. They were as follows :

Vehicle

Fuze

Type

Frequency

Bomb

T-82

Bar

White

Navy rocket

T-30

Ring

Brown

T-2005

Longitudinally

excited

Brown

Mortar shell

T-132

Longitudinally

excited

White

T-171

Longitudinally

excited

Brown

T-172

Transversely

excited

Yellow

5 61 Bomb Fuze T-82, Bar Type,

White Frequency

The bomb fuze T-82 was designed for the
same purposes as the M-166 (T-51-E1). When
it was found that the M-166 would meet mili-
tary requirements adequately and could be pro-
duced in quantity with a minimum of new
tooling, the need for the T-82 diminished and it
did not reach the production stage until just
before the end of World War II. The data given
here were obtained from pilot-production
samples and from the mass-production type-
approval sample.

The T-82 featured a turbine-driven generator
mounted in the base of the fuze. An air intake
port was provided through the center of the

fuze; there were two exit ports on opposite
sides of the fuze near the base (see Figure 28
of Chapter 4). The design had several advan-
tages over that of the other bomb fuzes. The
location of all moving parts close to the sup-
porting base, and well removed from those
sections of the electric circuit that are most
susceptible to mechanical disturbances, aided
in the production of a very stable fuze. On the
other hand, it was found that the variations in
turbine speed were somewhat greater than the
variations in propeller speed of the other bomb
fuzes. This appeared both as a greater spread
in air travel to arming of individual fuzes un-
der a given release condition and as a greater
dependence of air travel on bomb size.

Relative air travel on various bombs
M-30 M-81 M-64 M-65 M-66 M-56

1.00 1.02 1.24 1.48 2.32 1.87

Two models, the T-82-E1 set for 3,600 ft
MinSAT (not equipped with the arming delay
bracket) and the T-82-E2 set for 2,000 ft
MinSAT (equipped for arming delay device)
were current at the end of World War II. The
following data may be taken as representative
of the principal characteristics of both models.

The mechanical design of the T-82 is de-
scribed in Chapter 4; its principal components

Table 17. Pertinent features of T-82.

Electrical

Radio

Circuit

POD

+ 16.7

s(T f^lv 100)

\/p for R = °° /

Audio

PkAF

184

MvF (Pk)

MvF (140)

MvF (280)

C V (-volt)

7.7

EC (-volt)

3.4 (20 to 35K)

Proof performance*

Burst height

101

Reflection coefficient

0.8

Proper (%)

Random (%)

Dud (%)

Physical characteristics

Overall length (in.)

Length from shoulder (in.)

Overall width (in.)

Weight (lb)

Manufacturer:

Westinghouse

* From 10,000 ft at 200 mph on M-81 or M-88.

240

CATALOGUE OF FUZE TYPES

Figure 35. Power-oscillating-detector circuit
used in T-82 bar-type White-carrier bomb fuze.

are shown in Figure 28 of the same chapter.
The arming mechanism has to be built more
compactly than in the other bomb fuzes on
account of the lower position of the generator ;
otherwise it is essentially the same.

The power oscillating detector used in the
T-82 is shown in Figure 35. The component
values given in Table 18 are for the T-82-E2,
as required by Army Ordnance specification14
prepared in collaboration with Division 4.

Table 18. Component values of POD oscillator in
T-82 fuze.

Resistor

Value

(ohms)

Con-

denser

Value
(w*f) Coils

Turns

100,000

500 1

9 (antenna)

6.8

500

6 (plates)

(wound on same form)

6.8

Triodes

: NR3A

The nearly symmetrical oscillator assembly
(on a phenolic block) is shown in Figure 28,
Chapter 4. The two triodes are located on oppo-
site sides of the central air duct, in line with
the dipole bars. On a line at right angles are
the interwound plate and antenna coils (top)
and grid coil (bottom). The remaining oscilla-
tor components are disposed in a symmetrical
fashion with respect to these.

The basic amplifier circuit of the T-82 bar-
type bomb fuze is shown in Figure 32, Chapter

3. The component values given in Table 19 are
for the T-82-E2, as required by Army Ordnance
specification,14 prepared in collaboration with
Division 4.

Table 19. Component values for amplifier in
T-82-E2 fuze.

Resistor

Value

(megohms) Condenser

Value
(w*f )

2.2

10,000

0.3

115 ohms

3.3

R10

4.7

Rll

1.0

0.1 ^

R12

1.0

CIO

10,000

Cll

1,000

R14

330 ohms

C13

0.6

R15

1.0

C14

0.2 fii

R16

1.0

R17

30,000 ohms

R18

3,000 ohms

Pentode :

NS5

Notes. For R20, see R2 and R3 in oscillator circuit for the T-82.

A gain-control condenser Cg shown in Figure 16, Chapter 3, is
not present in the T-82 circuit, the gain of which may be adjusted
by selection of suitable value for Cll.

The transformer secondary is shunted by C3 and R6 in parallel
instead of in series, as appears in Figure 16, Chapter 3.

The general character of the amplifier assem-
bly of the T-82 differs somewhat from that of
the other bomb fuzes on account of space re-
quirements of the central air duct. The assem-
bly and its major parts are shown in Figure 13,
Chapter 6.

5 6 2 Navy Rocket Fuzes

T-30 (Mk 171, Mod 0) Ring Type,

Brown Carrier

The fuze T-30, like the T-2004, was a bomb
fuze modified for use on a Navy airborne
rocket. The T-30 was intended primarily for
attacking enemy aircraft with the HVAR. The
weakness of enemy opposition in the air during
the later stages of World War II made its pro-
duction less urgent than that of some of the
other fuzes. Although mass production (by
General Electric) had barely started when
World War II ended, considerable testing was
done with the pilot production model (includ-
ing a service test at the Naval Ordnance Test
Station, Inyokern) and its properties were
fairly well defined.

SECRET

PREPRODUCTION FUZES

241

Early functioning of the T-30 on account of
the considerable afterburning of the HVAR
propellant was a serious problem. The expedi-
ent of delaying arming until afterburning was
negligible and unsatisfactory because it gave
an undesirably large minimum firing range. A
program of research directed toward the elimi-
nation of afterburning was partially successful,
but the problem was by no means completely
solved at the end of World War II.

There was no fixed target testing with the
HVAR. Consequently, the data presented in
Table 20 for performance on this vehicle are
limited to high-angle firing tests.

Table 20. Pertinent features of T-30.

Electrical*

Radio

Circuit

+ 1

S (volt)

Audio

PkAF (c)

MvF (Pk)

CV (-volt)

7.8

EC (—volt)

3.9

Proof performance!

ROA (ft)

90 \

Proper + Mid (%)

77 (Vehicle: HVAR

Early (%)

20 ( QE: 55°

Dud (%)

3 )

Physical and arming characteristics ( same as T-2004)

Overall length (in.)

10^2

Length from shoulder (in.)

Overall width (in.)

Weight (lb)

* From General Electric units,
t Bowen units; no data available on GE.

The structure of the T-30 is practically the
same as that of the T-2004 and the OD-circuit
ring-type bomb fuzes and therefore requires
no additional description.

T-2005

The T-2005 was the logical next step after
the T-2004 and T-30 : a GP rocket fuze of small
size making use of the designs being developed
for mortar shell fuzes and provided with an
external switch to permit selection in the field
between two sets of characteristics appropriate
to plane-to-plane or plane-to-surface firing.
Test data are very scanty, and the characteris-
tics of this fuze can be represented best by the
tentative specifications that were completed
immediately before the end of World War II.

Some of the specification requirements are
shown in Table 21. The fuze is shown in Figure
46 of Chapter 4.

Table 21. Pertinent features of T-2005.

Plane- to-Plane

Plane- to-
Surface

Vane

Turbine

Same

Electrical

Radio

Circuit

RGD

Same

-2 to -8

Same

S (volt)

> 10 from 2K to

20K ohms

Same

Audio

PkAF

60 to 100

100 to 160

MvF (Pk)

15 to 30

70 to 130

MvF (75c)

115 to 105

CV (-volt)

6.8 to 8.5

Same

EC (-volt)

5.0 max

(15K to 50K rpm)

Same

Physical characteristics

Overall length

4f in.

Length from shoulder

4f in.

Overall width

2 \ in.

Weight

28 oz

The general design of the T-2005 (Figure 46,
Chapter 4) is similar to that of the T-171.
Since the ballistic effect of this fuze is less im-
portant than that of the mortar shell fuzes,
the antenna insulator is enlarged for increased
strength. The safety and arming requirements
placed on this fuze were rather complex, in-
volving a number of users’ options. The mech-
anism that was designed to meet these require-
ments is not readily described; reference is
made to Figure 47 and the accompanying text
in Chapter 4.

The electric circuit diagram of the T-2005
is shown in Figure 36.

3 6,3 Mortar Shell Fuzes

T-132, Longitudinally Excited, White Frequency

T-171, Longitudinally Excited, Brown Frequency

T-172, Transversely Excited, Yellow Frequency

All the mortar shell fuzes were considerably
smaller than the rocket and bomb fuzes. In
spite of this reduction in size, it was necessary
to use the larger tail of the M-56 shell when
they were mounted on the small M-43 mortar
shell in order to obtain stable flight. They were
designed primarily for use on 81-mm shells
such as the M-43 and the M-56.

The T-132 featured a radical innovation in
electric construction: the production of a con-

SECRET

242

CATALOGUE OF FUZE TYPES

siderable part of the electric circuit by painting
conducting material onto a ceramic base (see
Chapter 6). This technique was designed to
facilitate the maximum possible rate of pro-
duction.

Since immediate success of the new technique
could not be assured, the T-171 was developed
simultaneously, using standard components.

Another innovation was the loop antenna,
featured in the T-172.

The three fuzes are shown in Figure 6, Chap-
ter 1, and Figures 42, 43, and 44, Chapter 4.
Except for the items mentioned above, they
were quite similar in design.

None had entered mass production at the
close of World War II. Preparations for mass
production of the T-132 had been completed.
Considerable pilot production and develop-
mental testing data are available for this fuze.
Developmental testing data alone are available
for the T-171 and T-172. Some use is made of
specification requirements and design data in
representing the laboratory characteristics of
the latter two fuzes.

The field performance scores for the mortar
shell fuzes are of necessity averages over pe-
riods involving a number of design changes.
Although the scores are not impressive, they
compare favorably with those obtained with

Table 22. Pertinent features of T-132 (Globe-Union)

Arming (yd)

300 (approx)

Vane

Turbine

Electrical

Radio

Circuit

RGD

+ 11.6

S (volt)

(11 at 6,000 ohms)
( 9 at 20,000 ohms)

Audio

PkAF

107

MvF (Pk)

MvF (40)

CV (-volt)

7.5

EC (-volt)

5.0 max*

6.0 maxf

Proof performance*

Burst height (ft)

(over water)

Proper (%)

Random (%)

Dud (%)

Physical characteristics

Overall length

41 in.

Length from shoulder

3f in.

Overall width

2 in.

Weight

22 oz

* On raising generator speed from 20,000 to 80,000 rpm.
t On lowering generator speed from 80,000 to 0 rpm.
t On M-43 with M-56 tail charge: 1-4. QE: 45° to 80°.

SECRET

PREPRODUCTION FUZES

243

rocket and bomb fuzes at the same stage of
development. Pertinent features are shown in
Tables 22, 23, and 24. See Chapter 4 for further
structural details of the mortar shell fuzes.

The latest circuits of the mortar shell fuzes
are shown, with component values entered

Table 23. Pertinent features of T-171 (NBS).

Arming (yd)

300 (approx)

Vane

Turbine

Electrical

Radio

Circuit

RGD

—5 to +5

S (volt)

4 min (60 and 90K)

Audio

PkAF

80 to 120

MvF (Pk)

25 to 50

MvF (40)

60 to 140

CV (-volt)

6.6 to 8.7

5.0 max*

6.0 maxf

Proof performance]:

Burst height (ft)

(over water)

Proper (%)

Random (%)

Dud (%)

Physical characteristics

Overall length

4f in.

Length from shoulder

3f in.

Overall width

2 in.

Weight

22 oz

* On raising generator speed from 15,000 to 60,000 rpm.

t On lowering generator speed from

60,000 to 0 rpm.

X On M-43 with M-56 tail charge: 2

and 4. QE: 45°.

thereon, in Figures 37, 38, and 39. The circuits
for the T-132 and T-171 are taken from Army
Ordnance specifications,15’ 16 prepared in colla-
boration with Globe-Union, Inc., National
Bureau of Standards, and Division 4. The T-172
circuit is taken from the final progress report
of the Zenith Radio Corporation on this project.

The ceramic oscillator block of the T-132 in
various stages of “painting” of components and

Table 24. Pertinent features of T-172 (Zenith).

MinSAT (ft)

800

Vane

Turbine

Electrical

Radio

Circuit

RGD

+ 11

S (volt)

Audio

PkAF

Gain

Proof performance*

Burst height (ft)

(over water)

Proper (%)

Random (%)

Dud (%)

Physical characteristics

Overall length

6^ in.

Length from shoulder

5^ in.

Overall width of body

2 in.

Diameter of loop

3 in.

Weight

24 oz

* On M-43 with M-56 tail charge: 2, 3, 4. QE: 45° to 80°.

f SECRET

244

CATALOGUE OF FUZE TYPES

the non-painted components and the complete
assembly appear in Figures 10 and 7, Chapter
6. In the latter figure, the triode is seen to be
located in the center, in the position occupied by
the generator shaft in earlier fuzes; the small
white disks are the condensers.

The latest model of the T-132 amplifier is
shown in Figure 16, Chapter 6. The ceramic

plate is mounted parallel to the longitudinal
axis of the fuze. The figure shows both sides of
the plate, in both the “painted” state and in
the complete state. The reduction in size of this
assembly, relative to that of the earlier fuzes,
may be judged by comparing Figures 16 and
17, Chapter 6, noting that the electron tubes are
the same in both cases.

Figure 39.

Electric circuit of T-172 mortar shell fuze.

SECRET

INTRODUCTION

Chapter 6

PRODUCTION

6.1

IN the early days of radio proximity fuze
development, many workers in the field were
fearful that even though satisfactory models
might be built in the laboratory by skilled peo-
ple, the project would prove infeasible because
those models could not be mass-produced suc-
cessfully by unskilled labor in the huge quanti-
ties needed by the Services. Those who later
encountered and overcame some of the produc-
tion headaches that arose were, if the truth
were known, often of the same opinion. The
successive problems that arose were resolved
by the skill, perseverance, ingenuity, and op-
timism of the technical and production staffs
of the various manufacturers working in closest
cooperation with physicists and engineers of
the development staff.

It is the purpose of this chapter to outline
briefly some of the production procedures that
were adopted, some of the problems that arose
and their resolution, and, in general, to point
out some of the considerations involved in the
quantity production of generator-powered
proximity fuzes. No attempt will be made to go
into great detail regarding the manufacture of
any one type of fuze. Each type naturally has
its own peculiar production problems.

6,1,1 Pilot Plant Production

An effort was made to anticipate and over-
come the difficulties likely to be encountered in
mass production by means of pilot production
of considerable quantities of each type of fuze
in plants set up for that particular purpose.
These plants produced varying quantities of
preproduction models, in some cases as many

a This chapter was prepared by A. S. Clarke of the
Clarke Instrument Corporation, Silver Spring, Mary-
land, and consultant to the Ordnance Development Di-
vision of the National Bureau of Standards. Early in
World War II, he was technical aide to Division 4,
NDRC, and later manager of the Electronics Division,
Bowen & Company, Bethesda, Maryland, engaged in
pilot production of proximity fuzes.

as 25,000 of a given type. The conditions under
which these pilot plants operated were, to a
considerable degree, the same as would be en-
countered in large-scale production. The labor
was unskilled, or at best semiskilled; produc-
tion-line techniques were employed; and the
fuzes were not babied or hand-fitted at any
stage of manufacture.

These pilot lines served several important
functions in addition to developing production
procedures and establishing an index of pro-
duction feasibility and quality for the design.
They served as a source of fuzes necessary for
the extensive field testing of new designs, and
they provided a flexible source of supply for
fuzes modified from time to time as design
changes were dictated by field test results,
changes in tactical requirements, and by im-
provements by the design group.

6,1,2 Production Organization

To a first approximation, an organization for
mass production of proximity fuzes bears a
very close resemblance to the setup usually em-
ployed for mass production of radio receivers.
They are similar in that both organizations (1)
employ relatively unskilled labor, (2) break
down production into a multiplicity of simple
easily performed operations that can be quickly
taught to such labor, (3) make use to the fullest
possible extent of continuous production-line
methods, and (4) employ similar tools and
processes. The two organizations differ in that
in the fuze plant (1) more frequent and more
vigilant inspection is required, (2) closer liai-
son is required between engineering and pro-
duction departments, (3) more frequent test-
ing with more elaborate equipment is required,
(4) production supervisors should be of a
higher caliber since the emphasis must be on
quality of production rather than, primarily, on
quantity, and (5) eternal vigilance regarding
small and, in some other products, unimportant
details must be the rule.

245

246

PRODUCTION

If there is any formula for the successful
production of radio proximity fuzes, it would
probably read like this: Mix together equal
parts of careful workmanship, rigid inspection,
intelligent and responsible supervision, and
good production designs.

A process flow chart showing the routing of
fuzes through a typical plant is given in Fig-
ure 1.

Figure 1. Process chart for production of T-51
fuzes, Zenith Radio Corporation (reference 17,
Figure 3).

Typical interior views of plants engaged in
mass production of fuzes are given in Figures
2 and 3.

A typical plant layout is shown in Figure 4.

6,1,3 Preproduction Planning and
Preparation

One consideration that is fully appreciated
by every production man but often discounted
by design and laboratory workers is the neces-
sity for a truly tremendous amount of plan-
ning and preparation for quantity production

of even the simplest item. And the proximity
fuze is no simple item. The whole process of
mass production is a carefully integrated and
very delicate mechanism with its various parts
so interrelated and interdependent that a
breakdown at any one point can throw the

Figure 2. View of production line for T-50 fuze
at Emerson Radio and Phonograph Corporation
(Emerson photograph).

whole plant into complete disorder. Before pro-
duction justifying the name can begin, the
following must be done: (1) all drawings must
be completed, checked, and approved; (2) all
tooling, including production jigs and fixtures,

Figure 3. Another view of production line for
T-50 fuze at Emerson Radio and Phonograph
Corporation (Emerson photograph).

must have been fabricated and checked ; (3) all
test equipment must be completed, tested, and
installed; (4) supervisors must be trained;
(5) inspectors and test equipment operators

SECRET

OSCILLATOR

247

must be trained; and (6) an adequate supply of
every component or purchased subassembly
must be on hand to support a continuous flow of
production.

Of course there were occasions when, because
of the pressure of wartime urgencies, produc-
tion was started in advance of complete prepa-
ration as outlined above and some of the pro-
duction difficulties that arose are directly trace-
able to this situation.

62 OSCILLATOR

621 Introduction

The design of the oscillator portion of vari-
ous types of fuzes has been covered in consid-

The Problem

Basically, the problem is to mass-produce a
high-frequency oscillator of relatively small
size that will (1) feed adequate energy to a
suitable radiating system, (2) have the requi-
site frequency stability, (3) maintain the car-
rier frequency constant or within specified
limits from fuze to fuze, (4) give uniform out-
put from fuze to fuze, (5) be mechanically
rugged to withstand the shocks incident to
service, and (6) generate in itself no spurious
responses that might result in premature fuze
detonation.

All parts of the oscillator circuit of the fuze
are in strong r-f fields. Any motion of these
parts, either in relation to each other or to the
chassis, will produce a signal on the grid of the
amplifier tube that is indistinguishable from the

Figure 4. Plant layout for assembly lines for the production of T-51 fuzes, Zenith Radio Corporation
(reference 17, Figure 4).

erable detail in previous sections of this report.
No attempt will be made here to repeat this
descriptive material, and it is presumed that
the reader has studied and become familiar
with it. The purpose of this chapter is to outline
some of the procedures adopted in mass manu-
facture to insure that the requirements for a
satisfactory oscillator portion of the fuze are
fully met.

signal from the selected target that initiates
normal functioning. For this reason, every
effort must be made in production to produce a
rigid r-f assembly.

62 2 Typical Procedures

Usually the production department is handed
a layout and a model of the oscillator design.

SECRET

248

PRODUCTION

They are allowed little, if any, latitude in
changing the layout of components or the
method of oscillator construction. Even the
“dress” of every lead is specified, since very
slight variations in this respect might cause
undesirable variations in carrier frequency
from fuze to fuze.

Incoming Inspection

The first and a very important step in the
production of satisfactory oscillator assemblies
is adequate inspection of incoming components.
Those resistors and condensers used in parts
of the circuit having any effect on oscillator
drive, carrier frequency, or coupling to the
radiating system should be 100 per cent in-
spected. To what extent tubes should be in-
spected depends on the demonstrated reliability
of the inspection at the source of supply. Me-
chanical inspection should be made of the
coil forms for such defects as improper cur-
ing (poor mechanical strength), improperly
cleaned flash, and uniformity of size. The oscil-
lator mounting plate and tube shield assembly
should be checked for flatness, plating finish,
full complement of holes (small punches break
easily), and quality of soldering of tube shields
to the supporting plate. Molded oscillator blocks
and all other chassis parts should be inspected
for conformance to specifications. Mold pins
making small lead holes in plastic parts are
subject to frequent breakage and cases have
been known where a considerable quantity of
pieces have been molded and shipped before
such breakage was noticed.

Types of Oscillator Construction

Three different types of oscillator construc-
tion were in common use. Many of the mechani-
cal features of the designs have already been
discussed and illustrated in Chapters 3 and 4
(cf. Figures 13, 14, and 15, Chapter 3) of this
volume. For the purpose of discussion, the
types of construction listed below will be cov-
ered individually.

Basic types of oscillator construction

1. Phenolic block used for foundation.

2. Thermoplastic block for foundation.

3. Ceramic block used for foundation and in-
corporating so-called printed circuits.

Each type of construction presented its own
peculiar production problems. All of the blocks
had one common feature in that they made use
of molded cavities to support components and
employed cements of various kinds to anchor
these components in place.

Figures 5, 6, and 7 show oscillator assem-
blies employing the three types of construction
employed.

Thermosetting Phenolic Blocks

Where a mica-filled phenolic thermosetting
material is used for the oscillator foundation,
the first step is treatment to insure that all
moisture is driven from the block and that the
surfaces to which components are to be bonded
are made ready to receive the bonding agent.
Early production made use of a cement known
as Amphenol 912, a product of the American
Phenolic Corporation. This material does not
adhere well to the glazed surfaces of the phe-
nolic material as it comes from the mold. To
overcome this, the blocks are sand-blasted, a
treatment which also removes the glaze from
the sides of the coil cavities. If these sand-
blasted blocks remain very long in a humid
atmosphere, there is a possibility of additional
moisture absorption. To prevent this, after
sand-blasting the blocks are placed in a well-
ventilated oven or under infrared lamps and
heated to a temperature of approximately
150 F for a period of approximately three
hours, after which they are given a coat of
the 912 cement, which acts as a sealing agent
against further moisture absorption and also
furnishes a surface to which later applications
of the same cement would adhere more firmly.

It was found desirable by some producers to
treat also the larger components, such as tubes
and coil forms, with a light coat of the cement
and allow same to dry thoroughly before as-
sembly. This was found to aid materially the
later cementing of these components in place.

Later in the production program, cements
were found that bonded thoroughly with ther-
moplastic materials without the above sand-
blasting treatment.

Thermoplastic Oscillator Blocks

When blocks of thermoplastic material were

SECRET

OSCILLATOR

249

Figure 5. Oscillator assembly using thermosetting plastic block.

Figure 6. Oscillator assembly using thermoplastic block (Zenith photograph).

250

PRODUCTION

used, sand-blasting and precoating components
with cements was not necessary, since cements
were available that fuzed with the plastic ma-
terial to form a true bond.

The use of thermoplastic material for the
blocks had other decided advantages, one of
which was the ability to tack down leads along
the top surface of the block with a small hot-

through the use of a thermosetting cement.
Ceramic construction forms a special case
which is covered in more detail later in this
chapter.

Oscillator Coil Construction

In all fuze designs in use up to the end of
hostilities, the degree of carrier frequency uni-

Figure 7. Oscillator assembly using ceramic block (Globe-Union, Inc., photograph).

pointed metal tool which softened the material
sufficiently to allow the wires to be embedded
firmly into the block at that point.

Ceramic Blocks and “Printed” Circuits

The use of ceramic oscillator blocks with so-
called “printed” or “painted” circuits reduced
very materially, but did not entirely eliminate,
the number of loose components which had to
be anchored to the block. The interconnecting
wires and resistors were stenciled directly on
the ceramic and fired so that they became inte-
gral with the block. However, it was still neces-
sary to mount securely such items as coils,
tubes, r-f chokes, and condensers. This prob-
lem was successfully overcome by the only
manufacturer using this type of construction

formity that could be obtained was solely de-
pendent upon the uniformity of tube interelec-
trode capacities, the production of uniform
coils, and the reproducibility of wiring layout
and stray capacities from unit to unit. There
were no lumped capacities used in the fre-
quency-determining circuits. All the above ex-
cept uniformity of tube characteristics are
under the control of the fuze manufacturer. All
the factors influence not only the carrier fre-
quency but also the amount of oscillator drive
and coupling to the radiating element.

The uniformity of carrier frequencies and
oscillator output attained in production is
illustrated by Figures 8 and 9, the data for
which came from test on groups of approxi-
mately 500 units of each of three different man-

SECRET

OSCILLATOR

251

ufacturers. As a matter of interest, the degree
of carrier frequency uniformity achieved in
production sometimes proved embarrassingly
good, since, as a safety measure, some spread
in carrier frequencies is desirable.

The production of uniform coils is, in itself,
no mean achievement. In fuzes using a thermo-
setting mica-filled plastic for the oscillator
block, it was customary to use the same ma-
terial for the coil forms. These forms were
usually transfer-molded in multiple molds with

Coil leads were anchored in the above forms
by drawing the wire through accurately drilled
holes. These holes were drilled by inserting the
form in a drill jig made from a small square
block of steel having a close fitting hole for the
form and provision for uniquely orienting the
bosses on the end of the form with relation
to the holes to be drilled. Transverse holes were
then drilled through hardened drill bushings at

Figure 8. Uniformity of carrier frequencies in
production of radio proximity fuzes.

Figure 9. Uniformity of oscillator grid voltage
in large-scale production of radio proximity
fuzes.

removal from the mold accomplished by un-
screwing the piece from the cavity which was
essentially a tapped hole. A boss on the end of
the coil form served as a key for the wrench
used for removal. Another method used a two-
piece mold with the flash line occurring along
two flats on the side of the coil form. This latter
method has the disadvantage of requiring exact

the proper places. The mortality rate among
the No. 70 drills used for this purpose was very
high, due to the highly abrasive character of
the phenolic material.

All coils were wound by hand, using simple
winding fixtures in which the coil form was
turned by a crank which engaged the bosses on
the ends of the form. Before winding, the forms

252

PRODUCTION

were given a coat of Amphenol 912 cement and
dried. After winding, another coat of cement
was applied and allowed to dry thoroughly be-
fore the coil was inserted in its cavity and
cemented in place.

Coil forms molded of thermoplastic material
were a decided improvement in that it was pos-
sible to anchor the start and ending of the
winding by tacking the wire to the form
through the use of a small heated metal point.
In this manner it was possible to avoid some of
the troubles encountered in making uniform
coils in which the leads had to pass back
through the transverse holes drilled in the
form. In some cases, the wire emerged from the
form on the bottom or blind side of the coil and
had to be dressed back to the top of the chassis
block along the outside of the form, giving a
long and often indeterminately positioned lead
which sometimes resulted in variations in car-
rier frequency and loading and coupling. With
the thermoplastic forms, the actual end of the
helix of the coil could be located as desired on
the circumference of the coil, and while dou-
bling back of leads was sometimes necessary,
at least there was no wire threaded through an
oversized hole with the attendant worries as to
whether or not the lead could flop around in
the field of the coil.

On fuzes using interleaved and center-tapped
windings feeding transverse dipole antennas,
the center lead was usually formed by twisting
together an uncut loop in the continuous wind-
ing by means of an auxiliary fixture mounted at
right angles to the coil axis and having a crank-
actuated retractable button hook arrangement
around which was looped this center lead.
Turning the crank twists together the loop.
This twisted Formvar pair of coated wire then
had the insulation removed by dipping it in
what was essentially a superheated solder pot.
This pot was a metal tube approximately 4 in.
long and % in. in diameter surrounded by an
electric heating element which kept the melted
solder at a temperature of approximately
1200 F. Since the pot diameter was small, only
a very little area of the solder pool was exposed
to oxidation. At the temperature used, the
Formvar insulation immediately melted off and
wires became well tinned and intimately con-

nected as a single lead. This lead could then be
cut to the proper length without unraveling.

The production of coils wound on ceramic
forms is covered in a special Section 6.2.3.

Mounting of Tubes in Oscillator Assembly

One of the oscillator components most diffi-
cult to mount securely in the assembly is the
tube, or tubes if a diode is also used. Many ex-
pedients were devised for this purpose; some
of the more successful ones are described. As
can be seen in Figures 5, 6, and 7 (also Figure
8, Chapter 3), the tubes used for oscillators
were, in the majority of cases, oval in shape,
having two relatively broad flat sides with nar-
row rounded ends. The glass seal-off tip on the
end of the tube was the only portion that
approached the rounded bottom of the tube
well, and at best the tube was in contact with
the round metal shield at only three points.
The problem then is how to overcome this
basically weak mechanical design.

By far the majority of early fuze production
had the tubes cemented in place. Toward the
end of the program, other and better methods
had come into use. In cementing in tubes, using
the aforementioned Amphenol cement, there
was always the possibility of getting too much
cement in the tube shield and having it “case-
harden.” This name is given to a particularly
undesirable condition that arises when a hard
film forms on the surface of the pool of cement
during rapid initial evaporation of the solvent.
This hard film then serves as a trap which
prevents the further evaporation of the re-
maining solvent, which in the course of time
permeates and softens that portion of the
previously dried cement that was presumably
holding the tube in place. The same condition
also prevailed in cementing coils in the cavities
provided for them in the oscillator block.

One method of preventing the above difficulty
and of insuring a fairly uniform application of
cement all over the block was evolved and
promptly dubbed the “fruit jar” technique.
Here the cement of carefully controlled vis-
cosity was contained in a round-mouthed jar
and the oscillator assembly placed top down
across the mouth of the jar. The whole was
then inverted so that copious quantities of the

OSCILLATOR

253

cement permeated every cavity on the block and
completely surrounded all components, includ-
ing the tube. The jar was then returned to the
normal position and the surplus cement allowed
to drain back into the jar. Sufficient cement was
retained around the components in the cavities
and tube wells through capillary attraction to
form well-defined fillets which, since they con-
tained a minimum volume of cement, dried
hard quickly and adequately held the compo-
nents.

When the above cement dunking process
was combined with shaped tube shields, the
ultimate obtainable with the cementing method
of tube placement resulted. Shields were shaped
by inserting a mandrel having approximately
the same size and shape of the tube into the
normally round tube shield and squeezing the
shield around it. The result is a shaped well
which is a neat push fit for the tube with
several points of contact between the tube
and the shield. Even where there is no con-
tact, the separation is so small that cement was
retained in the space after the dunking process.

One manufacturer used glass wool wrapped
around the tube before insertion in the shield.
This wool served to cushion the tube slightly
and its compressibility permitted a neat wedge
fit of the tube in the shield. Somewhat the same
effect was obtained by another manufacturer
who used a rubber cushion around the tube.

Toward the end of the production program,
a tube potting compound was evolved consist-
ing of 80 per cent microcrystalline wax and 20
per cent polyisobutylene. This molten mixture
sets up rapidly upon cooling, providing a secure
anchor for the tube. The technique is fast and
simple and lends itself admirably to rapid pro-
duction. It was in use by at least two manufac-
turers.

6'2'3 Production of Ceramic

Oscillator Assemblies

As mentioned previously in this report, the
T-132 fuze employed unique methods of con-
struction based on the use of so-called “printed”
circuits on steatite plates and blocks. Because
of its advantages and potentialities for future

use, it seems advisable to present as much in-
formation as is available on this type of fuze
construction in this special section of this chap-
ter. All the information is from a report by the
Globe-Union Company to Division 4, NDRC, on
development work performed under an OSRD
contract on development work on this type of
construction.8

This design is built around the process of
forming the interconnecting leads and resistors
for the circuit by the application of conducting
and semiconducting materials directly to the
surface of molded ceramic plates or blocks and
the subsequent baking or firing of such ma-
terials to the extent that they form virtually
an integral part of the block itself. The tech-
niques used for metalizing and resistoring are
considered by Globe-Union to be trade secrets,
although they state in the same report that the
general methods are well known to the art. It
is important to realize that the end result may
be obtained by different manufacturing proc-
esses, and it is not essential that the identical
processes and techniques employed by Globe-
Union be used. The metalizing art is an old
development of ceramic and glass industries
and there are many widely used methods of
metalizing in use by the industry. The resistor-
ing process incorporates the processes devel-
oped in the making of variable resistors and
this too is a widely known art.

Construction of Ceramic Oscillator Block

The oscillator block is shown in Figure 7. The
steatite used has passed the Army and Navy
qualifications tests in accordance with JAN
Specification 1-10 and is known as a grade L-5
ceramic. The properties of the material are
tabulated below.

Mechanical

Specific gravity
Modulus of rupture
Tensile strength
Compressive strength
Coefficient of thermal
sion

Moisture absorption
Impact strength

Power factor
Dielectric constant
Loss factor
Dielectric strength

2.5

20,500 psi
9,100 psi

76.000 psi

expan-

6.9 X 10-6 20 to 100 C
Less than 0.02 per cent

2.0 ft-lb per in.
Electrical

0.110 per cent (1 me)
5.82 (1 me)

0.640 per cent (1 me)
247 v per mil

Secret

254

PRODUCTION

Because of the large and intricate shape of
the block, it was molded by the wet process to
provide better flow characteristics in the mold.
Allowance for shrinkage during firing was
made. Figure 10 shows the block as it comes
from the mold. At this stage, it is completely
formed except for the outside dimensions and
the two large coil holes. The coil holes were

the ceramic, a thin film of silver was applied to
the ceramic as described below. Additional
coats of copper, tin, and solder were applied to
the base coat of silver on certain parts of the
block where it was desired to reduce the electric
resistance of the coating to a very low value or
to facilitate soldering to heavy metal parts,
such as the support member and the shell.

Figure 10. Ceramic oscillator block in advanced stages of preparation. View in upper left shows block
as it comes from mold. (Globe-Union, Inc., photograph.)

drilled and the outside grooves and proper di-
mensions were obtained by machining opera-
tions. The block is then fired in a kiln to pro-
duce a hard, white, vitrified material.

Construction of Ceramic Coil Forms

The oscillator, antenna, and choke coil forms
were extruded in the form of rod from the same
type of material used for the oscillator block.
All three forms were fluted to facilitate wind-
ing and both antenna and oscillator coil forms
are threaded for accurate spacing of the wind-
ing. Threads and lead holes were machine cut
in separate operations before the forms were
fired.

Metalizing of Oscillator Block

To provide circuit connections and means of
fastening other metal parts to the surface of

The material used for silvering consisted of
finely divided silver powder in a suitable ve-
hicle. Before application, the surface of the
ceramic was thoroughly cleaned to remove all
trace of oil and dirt. After it was applied, the
piece was fired in an oven to burn out the
vehicle and cause the individual particles to
coalesce, forming a continuous film which ad-
heres tenaciously to the surface of the ceramic.

Owing to the irregular shape of the oscillator
block, the silvering material was applied by a
roll to the edge and by brush to the circuit ele-
ments. The edges were silvered to enable the
shell and support member to be soldered di-
rectly to the block. A tracing template was used
to position accurately the location of the sur-
face connections and a small brush was used to
apply the material to the surface.

Both edges were given a plating of copper

OSCILLATOR

255

and tin over the initial silver coat to facilitate
soldering. All high-current leads were copper-
plated to provide low-resistance paths. The
average thickness of the silver coating was
approximately 0.0002 in.

Resistoring Process

The process of resistoring on the ceramic
surface consisted of applying a suitable resist-
ance material between two metalized elements
on the surface. The resistor material consisted
of a base of conducting particles, such as car-
bon black or graphite, in a suitable vehicle. The
surface to be coated was covered with a mask
having suitable cutouts to outline the areas on
which the material was to be deposited and the
material was applied to the surface by spray-
ing. After spraying, the resistor was air dried,
the mask removed, and the resistor baked to
stabilize the resistance. After baking, the re-
sistor was checked for value, and if any adjust-
ment was needed it was made by scraping
away a little of the resistance material to in-
crease the resistance value. This was possible,
since in the spraying operation the low side
of the resistance tolerance was favored. Once
the resistor was adjusted, it was given a pro-
tective coating of varnish.

A number of factors are important in deter-
mining the resistance value. The variable fac-
tors are the ratio of the conducting particles to
the vehicle or binder, and the length, width,
and thickness of the deposit. The air pressure
used in the spray gun and the baking time were
found to have no appreciable effect on the re-
sistance value.

Resistors made as above described exhibit a
slight negative voltage characteristic, as shown
in Figure 11, and have good stability under
adverse humidity conditions. Tests have indi-
cated that they will dissipate approximately
0.3 to 0.4 watt for a period of 250 hours with a
decrease in resistance of only 7 per cent.

Soldering to the Ceramic Surfaces

Special techniques for soldering to the ce-
ramic were used in order that the following
requirements could be met: (1) that the solder-
ing process not weaken the ceramic because of
heat shock, (2) that initial strains not be de-

veloped in the ceramic because of excessive
shrinkage of the solder upon cooling, and (3)
that the solder not dissolve the thin film of
metal on the surface of the ceramic. The use of
special low-temperature low-contraction solder
such as RM 275, together with preheating of
the ceramic, prevents heat shock from occur-
ring. For soldering directly to the silver coat-
ing, a special low-temperature silver alloy
solder such as RM 297 was used. Use of silver

Figure 11. Variation of resistance with voltage
for “painted” resistors used in ceramic assem-
blies.

alloy solder prevents the silver deposit on the
surface from being dissolved into the solder,
since the solder is already saturated with silver.

Ceramic disk capacitors and wire leads were
soldered directly to the ceramic using the ap-
propriate solder. The resulting bond between
metal and ceramic was very strong, and it is
possible to rupture the ceramic before the joint
will fail.

Ceramic disk condensers are soldered directly
to appropriately located silvered points on the
plates. This was very simply done by heating
the disk capacitor to a temperature sufficient
to melt the solder, applying a small amount to
one face and pressing it against the ceramic
plate with sufficient heat to cause the solder to
bond to the silvered surface of the plate. Con-
nection to the top side of the condenser is made
by soldering a small strip of metal ribbon to

SECRET

256

PRODUCTION

that surface, the ribbon then being connected
to any desired point of the circuit.

The soldered joints between the oscillator
block and the shell and support member are
very critical, since the even distribution of the
weight of the unit to the oscillator block is de-
pendent upon the quality of these joints. The
block was grooved at these two joints to pro-
vide a capillary trough to insure that there was
a secure bond between the inner face of the
shell and support member to the block. In
soldering these joints, the operator was re-
quired to use a precut specified amount of
solder to insure that the joints were completely
filled.

Assembly of the Ceramic Oscillator Block

Assembly of the tubes, coils, and chokes to
the oscillator block presented a mechanical
problem due to the axial mounting of these
parts and the necessity of their remaining in
fixed position during setback. Coils, chokes,
and tubes were first held in place with polysty-
rene cement. This cement was chosen because
of its excellent high-frequency dielectric prop-
erties, but its use necessitated long and careful
drying under infrared lamps. The tube was
held in place in the oscillator block tube well
by wrapping the base of the tube with glass
wool and impregnating the wool with polysty-
rene cement. This made a large mass wetted
with the cement and which dried very slowly
even under the application of considerable heat.
A film would form over the surface, preventing
the rapid evaporation of the remaining solvent.
In many cases, tubes thus cemented slid out of
position during setback, and it was found nec-
essary to devise another method of holding the
tube. This was accomplished by cementing the
tube in place with a thermosetting cement. The
tube was positioned by a jig and the tube well
filled with cement up to the level of the leads.
In order that the cementing time required for
the blocks be reduced to the minimum and the
long drying cycle previously necessary for the
polystyrene cement be eliminated, the coils and
chokes were also cemented in place with the
thermosetting cement. A subsequent baking
cycle of 3 hours hardened the cement.

All oscillator block assemblies were ad-

justed to draw a total plate current of a speci-
fied value by the addition of a padder resistor
of conventional construction mounted axially
in the block. After cementing, the necessary
value of resistance was determined and the
proper resistor attached to the block. Polysty-
rene cement was used to anchor this resistor
in place; however, due to the heavy leads by
which it was attached, the resistor was self-
supporting and thorough drying of the cement
before further assembly work was not found
necessary.

The assembled oscillator block, together with
all the components entering into the assembly,
.is shown in Figure 7.

AMPLIFIERS

6,3-1 Requirements

The essential characteristics of amplifier de-
signs for satisfactory fuze operation have been
covered in Section 3.2 of this volume. This pre-
vious discussion does not deal with methods of
construction nor with the various processes
and procedures used in mass production of am-
plifiers having the desired electric character-
istics.

The production department is usually handed
a circuit diagram, a model of the type of con-
struction proposed by the engineering depart-
ment, a set of specifications covering perform-
ance of the finished unit, and a list of the pre-
cautions and procedures found necessary by
the engineering department in their model
work on that particular design. From this point
on, it is the responsibility of the production
department to mass-produce amplifiers having
the desired characteristics and meeting the
stated specifications with the least possible
assistance from the engineering and develop-
ment group.

The two basic requirements of the amplifier
are that it have the desired gain and shaping.
At this point it should be pointed out that the
term “gain” as used in connection with fuze
production does not have quite the same conno-
tation as it ordinarily possesses. The gain of
the amplifier system, as such, is seldom meas-

SECRET

AMPLIFIERS

257

ured in production. What is actually measured
is the overall figure of merit known as “milli-
volts to fire.” This takes into account the effec-
tive critical voltage of the thyratron and the
amplitude of hum (generator ripple) and
spurious voltages originating in the amplifier
circuit.

The necessity for shaping the amplifier gain
characteristic for the desired frequency re-
sponse has been thoroughly covered in Section
3.2. Examples of the shaping necessary to meet
the basic requirements have been shown and
methods of obtaining such shaping discussed.

the amplifier which critically affect the gain
and shaping should be 100 per cent inspected.

Obviously, the type of circuit employed in-
fluences the mechanical layout and construction
of the amplifier. The four principal types of
construction employed in fuzes that reached
the production stage were as follows: (1)
sandwich or wafer, (2) ring or collar, (3)
printed circuits on ceramic plates, and (4) disk.

Sandwich Construction

A typical amplifier using the sandwich type
of construction is shown in Figure 12. In this

Figure 12. Sandwich-type assembly for amplifier.

Procedures

It is the purpose of this section of the chap-
ter to deal with types of amplifier construction
and the various manufacturing procedures em-
ployed.

As with the oscillator, the construction of
satisfactory amplifiers begins with inspection
of incoming components. Those components of

construction, two punched linen Bakelite plates
approximately %2 in. in thickness are held
apart in a suitable jig and most of the resistors
threaded through holes in the plates. The two
plates are then pushed together with the re-
sistors acting as spacers. This foundation then
passes down the production line and has pro-
gressively added to it other resistors, con-
densers, and tubes.

SECRET

258

PRODUCTION

This type of construction has certain advan-
tages and, as usual, certain disadvantages. The
construction results in a rigid assembly and
shorter leads than is possible with any other
method using conventional components. It has
one decided disadvantage in that it is not pos-
sible to replace a defective resistor after the

One variation of the sandwich-type construc-
tion is shown in Figure 13. This design was
evolved because of the need for an unobstructed
passage for air to operate an internal turbine
(T-82 fuze). Two round Bakelite disks having
a central hole for the air tube are utilized. On
the two disks are mounted the electric com-

Figure 13. Sandwich-type assembly for amplifier with central opening.

sandwich is put together. In practice, however,
this has not proved as much a drawback as it
might seem, since very few amplifiers get
through with defective or incorrect resistors in
place.

ponents with some of the larger components
sandwiched in between. This type of construc-
tion was used by one manufacturer only (West-
inghouse), and large-scale production was just
beginning at the end of hostilities.

AMPLIFIERS

259

Ring Construction

The ring or “dog collar” type of construction
found considerable favor with production peo-
ple. The majority of the fuzes manufactured
(particularly T-51) used this type, a typical
example of which is shown in Figure 14. In this
construction, a strip of Bakelite or fish paper
is punched to receive either eyelets or, in some
cases, lugs for attachment and interconnection
of various resistors and condensers which are
attached to one end and, in the case of some
manufacturers, both sides of the strip as it
progresses down the production line. At some
point in the line, the strip is bent into a circle

Several different types of amplifiers were de-
signed making use of this process. One, shown
in Figure 15, was essentially a sandwich
composed of two horizontally mounted ceramic
plates containing “printed” resistors and inter-
connections, between which were mounted the
larger paper condensers and tubes. This type
of construction was abandoned because of me-
chanical weakness and replaced by a single
ceramic plate mounted on edge for greater
resistance to breakage under setback condi-
tions. This amplifier is shown in Figure 16.
This illustration shows the ceramic plate am-
plifier in various stages of production. At the

Figure 14. Ring-type assembly for amplifier (Zenith photograph) .

and the two ends riveted together. This ring,
or collar, is then inserted in the amplifier
cavity. Because of its shape, this design prob-
ably makes for maximum utilization of the
space available. All components are accessible,
and amplifiers having defective components or
incorrect values installed can be readily sal-
vaged.

Ceramic Amplifiers

The “printed” circuit on ceramic plates car-
ries out the same methods of construction as
were discussed in some detail in Section 6.2.1.

right is the plate after all silvered interconnec-
tion jumpers have been fired on. Second from
the right shows the plate with all resistors in
place, while the two views to the left show
opposite sides of the finished amplifier with all
paper and ceramic condensers and tubes in
place. The methods of applying resistors and
interconnection leads on these plates was iden-
tical to those described in the portion of this
chapter covering oscillators (see Section 6.2.3).

Disk Construction

This disk construction, which is illustrated in

SECRET

260

PRODUCTION

Figure 15. Components and assembly of ceramic-type amplifier, early version (Globe-Union, Inc.,
photograph).

Figure 16. Components and assembly of ceramic-type amplifier, late version (Globe-Union, Inc.,
photograph) .

AMPLIFIERS

261

Figure 17, is, in reality, a variation of the
sandwich previously discussed. The view shows
two types of disk construction as compared
with a conventional sandwich assembly shown
in the center. The components are laid flat
against the upper and lower surfaces of two
suitably punched Bakelite strips and wired up,
the two plates later being interconnected to
form the amplifier assembly. The construction
provides for maximum accessibility during
fabrication and was particularly favored by
one manufacturer.

Gain-Control Condensers

The methods of construction used in gain-
control condensers is of considerable interest.
These small capacitors, which are used to ad-

change is made by peeling off more or less of
the wrapped wire. This provides a method for
changing the capacity in very small increments.
One of the finished condensers of this type is
shown just below the tubes in Figure 12. In
fuzes using the ceramic plate construction, it
was the practice to determine the amount of
capacity needed by means of a continuously
variable condenser, which is part of the ampli-
fier test fixture, and then to select from previ-
ously graded groups of condensers the indi-
cated size of fixed capacity, which was then
wired permanently into the circuit. This
method consumed about the same amount of
time as the adjustment of a gimmick wire and
has the important advantage that it makes for
greater stability in the amplifier and less

Figure 17. Disk-type of amplifier assembly showing comparison with sandwich type. Latter is in center.

just the gain of the amplifier by control of re-
generation, vary in capacitance from approxi-
mately 2 to 30 \i\if. The majority of manufac-
turers used a modification of what is termed a
gimmick in radio receiver manufacturing par-
lance. In this device, the capacity is formed be-
tween a piece of enameled copper wire, usually
about size 18, acting as a mandrel, and a piece
of smaller diameter enameled wire wrapped
tightly around it. The enamel insulation on the
two wires forms the dielectric, and capacity

change of amplifier characteristics with pot-
ting. Attention is called to the precautions nec-
essary to protect the gimmick type of condenser
from changes in capacity due to the potting
materials and process which are described in
the next section.

Potting and Impregnating Procedures

It was required that all amplifier assemblies
be embedded in a potting compound in order
that the electric characteristics would remain

EGRET

262

PRODUCTION

stable throughout the various conditions of
storage and use. Usually, some preliminary im-
pregnating processes were necessary before
final potting (see Section 4.7.6).

In amplifiers using Bakelite or fish-paper
strips as the foundation, precautions are nec-
essary to prevent these materials from absorb-
ing atmospheric moisture, which might result
in relatively low-impedance paths across criti-

40,

z 40. ,

Figure 18. Uniformity of frequency of peak
audio amplification in large-scale production of
radio proximity fuzes.

cal portions of the circuit and adversely affect
the operation of the amplifier.

One manufacturer’s procedure involved the
immersion of the fully punched amplifier ter-
minal strip and also the insulator strip, used
to prevent shorting of components to the metal
case, in hot ceresin wax until all bubbling
stopped. All the amplifier parts and tubes are
then mounted and the entire assembly again
impregnated with hot ceresin. After cooling,
it is then flash-dipped so that a heavy protec-
tive layer of wax is deposited on all parts.
These treatments serve to drive out and keep

out moisture and at the same time prevent any
appreciable, if not all, deleterious effects from
the tung oil potting compound.

Essentially the same procedure was followed
by other manufacturers, about the only varia-
tion being that instead of ceresin, some manu-
facturers used commercial microcrystalline
waxes sold under such trade names as Superla
or Halowax.

One interesting variation of the above pro-
cedure was used in pilot production of T-30
fuzes. This procedure might prove awkward in
large-scale operations because of the larger
quantities of assemblies involved. In order to
drive out all moisture before impregnation, the
day’s production of completed amplifier assem-
blies (before installation of the gain-control
condenser) were accumulated and placed on top
of cold hard Superla wax contained in pans.
These were then placed in an oven, together
with an active drying agent, and baked at about
70 C for 8 hours. During the first 4 hours, the
wax does not become sufficiently molten to
allow the amplifier assemblies to sink below the
surface. Thus, the amplifiers were actually
baked in a drying atmosphere before impreg-
nation. After the wax completely melts, the
amplifiers sink and are cooked for the next 4
hours in the hot wax. Probably the only way to
improve on this process is to vacuum impreg-
nate the amplifiers, but this procedure is some-
what awkward where wax is used as the im-
pregnating agent.

The gimmick-type gain-control condensers
must be protected against the action of the tung
oil potting material. Many schemes were tried,
the most successful being the boiling of the
finished condenser assemblies, with the ends of
the wound wire twisted together, in Zophar
Mills No. 1563 Wax at about 150 C for 4 hours
to drive out air and fill all cavities with the
wax. The condensers are then removed, the
free ends of the outside wire clipped short, and
the condensers boiled for another 4 hours in
order to allow the winding to assume a relaxed
or normalized condition. This tends to avoid the
effect of further unwinding and the resulting
change of capacity after adjustment of the gain
of the amplifier.

After the above treatment, the gain-control

SECRET

AMPLIFIERS

263

condensers were inserted in amplifiers previ-
ously impregnated as described and gain ad-
justment and final check made on the amplifier.
The accepted amplifiers were then flash-dipped
again in Super la wax at about 75 C to seal off
the clipped end of the gimmick wire against
moisture absorption.

It is interesting to note the degree of uni-
formity obtained by various manufacturers in

facturer at the Central Testing Laboratory at
the National Bureau of Standards.

Before potting amplifiers in the fuze cavities,
it is desirable to preheat the fuze by baking in
an oven for about 1 hour at 45 C. This baking
process accomplishes two results: (1) it dries
out the amplifier cavity, and (2) it provides
a warm surface for contact with the tung oil.
This hastens the polymerization and minimizes

Z 111

3 I

Figure 19. Uniformity of millivolts to fire in mass production of three different types of radio proximity
fuzes: A, maximum and minimum values of millivolts to fire shown throughout broad pass band of T-51
amplifier: B, spread in peak millivolts to fire is shown for narrow pass-band amplifier of T-90 (top) and
T-89 (bottom).

holding the peak audio frequency and milli-
volts to fire at peak audio frequency to the de-
sired limits. The spread of the peak audio-fre-
quency values around the design center for
three different manufacturers are shown in
Figure 18. Figure 19 shows the spread of milli-
volts to fire at peak frequency around the de-
sign center for three different manufacturers.
These figures are based on an analysis of tests
made on approximately 500 units of each manu-

the time during which the active tung oil mix
can attack the wax on the amplifier assembly.
Wax is a better insulator than tung oil, so that
for the purposes of amplifier uniformity, re-
moval of the wax coating must be prevented.
If quick polymerization is effected, less harm
is done to the wax.

The proportions of tung oil and polymerizer
used by different manufacturers varied from 5
parts of tung oil to 1 part polymerizer to as

ECRET

264

PRODUCTION

high as 15 to 1. Since the polymerizer is slightly
corrosive, there is some advantage in using as
little of it as possible. The high ratio of tung oil
to “hardener,” however, does make the setting-
up time of the material longer.

The polymerizer, or hardener as it was usu-
ally called, was available from Westinghouse
as an already prepared material, and most
quantity manufacturers used this source of
supply. Instructions for the preparation of this
hardener are included here as a matter of rec-
ord. The quantities given are for 1 gallon of

Figure 20. Vacuum fixture for potting ampli-
fier units (Globe-Union, Inc., photograph).

hardener: ferric chloride, 6.4 oz by weight;
tri-cresyl phosphate, 1 lb 2 oz by weight, castor
oil, 3 pt, 7 fluid oz.

Great care must be given to the handling of
the hardener ingredients, particularly the an-
hydrous ferric chloride, in order to guard
against contamination with moist air. The an-
hydrous ferric chloride is added slowly to the
tri-cresyl phosphate, stirring constantly with
a motor-driven stirrer for 2 hours. This should
be done in a narrow-mouthed container to re-
duce the circulation of air over the exposed
surface and the amount of surface exposed. The
castor oil is then added and stirred until thor-
oughly mixed. The hardener is then poured
into sealed containers.

The hardener and tung oil are combined and
thoroughly mixed by a motor-driven stirrer in
a covered container for approximately 5 min-
utes, after which it is poured into the amplifier
cavities. Owing to the viscosity of the mix, sev-
eral intermediate pourings are usually required

as the level gradually settles. After pouring, the
fuzes are then returned to an oven held at
approximately 45 C and kept there for approxi-
mately an hour to hasten polymerization.

Different methods of setting up the potting
operation as an integral part of the production
line were devised by different manufacturers.
In one plant, the conveyor belt was routed by
an air-conditioned room in which all mixing
and pouring operations were conducted. The
units were placed on the conveyor belt and car-
ried through the preheating oven. As they
passed a window of the mixing and pouring
room, the rack containing a group of units was
pulled through the small opening into the pot-
ting room where the units were filled. The rack
was then placed back on the belt and the units
continued on through the oven for a sufficient
length of time to permit setting up of the ma-
terial.

In another plant, the materials were mixed
in large refrigerated containers and dispensed
from this central point to a number of small
containers on the assembly line, also refriger-
ated, and the units filled by gravity flow from
these secondary containers.

Large refrigerated tanks were used on the
line of another manufacturer, each holding
approximately 15 gallons of the potting mix-
ture. These tanks were equipped with motor-
driven agitators to keep the mixture continu-
ally stirred to prevent separation or stratifica-
tion of the hardener and tung oil. Air was
applied under pressure to the top of these tanks
and the mixture was consequently ejected rap-
idly into the fuze cavity through flexible plastic
tubes.

One manufacturer used a vacuum potting
process which is of some interest. The fixture
used is shown in Figure 20. The glass tubes
were filled with the tung oil mix to the height
marked on the tubes. The units were placed in
a vacuum tank made of heavy plate glass. After
the desired degree of vacuum had been drawn,
the stopcocks were opened and the liquid flowed
rapidly into the fuze directly underneath the
tube. It was claimed by the manufacturer that
this method of potting resulted in better pene-
tration of the potting compound into the voids
in the fuze cavity.

NOSE ASSEMBLY

265

In addition to the tung oil mixture described
above, two manufacturers used what was
known as “Glidden” potting compound made by
the Glidden Company, of Cleveland, Ohio. This
material is a mixture of linseed oil, fatty acids,
rosin, magnesium oxide, and alkaline-washed
linseed oil. Its use requires the same careful
temperature control as tung oil to prevent pre-
mature setting up. It is slightly more difficult to
pour and is not quite as good mechanically as
tung oil, but it has a very decided advantage
over tung oil in that it is not as corrosive. One

Figure 21. Arbor press for staking windmill
bearing assemblies.

manufacturer using Glidden compound very
materially reduced the percentage of amplifiers
rejected for change in sensitivity after potting.

6 4 NOSE ASSEMBLY

Other chapters of this report have covered
the evolution of a satisfactory design for the
vane bearing and rotating system used on the
majority of fuzes. Attention is particularly
called to Section 4.3.2.

Production difficulties with the nose assem-
bly centered principally around the problem of

obtaining satisfactory bearings for the vane
shaft. Early attempts to use porous bronze
sleeve bearings proved unsatisfactory because
of the high rotational speeds encountered in
service. Figure 15 of Chapter 4 shows the type
of bearing used on the first fuzes placed in pro-
duction. Figure 18B of the same chapter shows
the bearing in cross section. As can be seen in
these figures, the nose bearing somewhat re-
sembled a bicycle wheel bearing. A steel sleeve
bearing having recesses at both ends was
molded into the plastic nose. Staked to the
metal propeller or molded as an insert in the
plastic propeller was a shaft having a hardened
conical surface at the vane end and a thread on
the other end. On this threaded portion was
placed a nut having a hardened conical surface
similar to the one on the vane shaft. The bear-
ing surfaces of the shaft and nut were selec-
tively hardened by stopping off the unhardened
portion by copper plating. This plating in-
hibited the action of the cyanide case harden-
ing solution. Steel balls were placed in the re-
cessed ends of the bearing sleeve and contacted
two sides of the recess and the above-mentioned
conical surfaces.

In production, simple fixtures were used to
place a predetermined number of balls in each
race, and the nut was tightened up by hand
until the feel of the bearing was slightly looser
than the desired end condition. The assembly
was then placed upside down in a staking fix-
ture, shown in Figure 21. This fixture was
built from a conventional arbor press and
serves to guide a hardened tool having two
sharp projections down into the slotted vane
lock nut, where these projections shear and
force two tabs of metal from the shaft into a
smaller transverse slot in the nut, thus keying
the shaft securely to the nut and acting as a
means for transferring torque and preventing
backing off or loosening of the nut. Since there
was inevitably some play between the threads
on the shaft and the nut, this staking operation
also forced the nut farther down on the shaft
until all play in the threads was eliminated.

The whole trick of this staking operation was
to get the desired degree of tightness or pre-
loading for the bearing without indenting the
soft races in the steel insert sleeve. On the side

266

PRODUCTION

of the ram (see Figure 21) is an eccentric stop
nut which limits the downward motion of the
ram and consequently the pressure applied by
the spring to the staking tool. On the first try,
this stop nut was set at some arbitrary mini-
mum position and the ram advanced until lim-
ited by the stop. The nose assembly was then
removed from the fixture and the operator
judged the feel of the bearings by manually ro-
tating the propeller. Based upon experience,
this feel gave an indication of how the stop nut
should be adjusted for the next stroke. By this
method, a satisfactory bearing was usually
obtained in not over three adjustments, with
rejects due to overshooting the desired pres-
sure not over 4 per cent. A reasonably intelli-
gent operator could be trained for this opera-
tion in a day.

Associated with the above bearing design was
a coupling shaft whose limitations have been
outlined in Chapter 4. The design of the rotating
system was probably the best possible in view
of the necessity of using something other than
commercial ball-bearings, which were not
available in the quantities needed for the pro-
gram at the time fuze production was started.
When commercial bearings became available,
it was possible to change to a design which is
illustrated in Figure 16, Chapter 4. In this de-
sign, the shaft extending back to the generator
was integral with the vane. There was enough
play in the commercial bearing to allow for
small angular misalignment between the gen-
erator shaft and the nose. The bearing was
dropped into place in the recess provided in
the metal insert molded in the plastic nose and
held in place by either staking or rolling over
the edge of the recess. There was no fitting of
bearings or variation in the tightness of the
bearings caused by human judgment.

It was necessary, in order to reduce vibra-
tion, to balance the vane dynamically. The
equipment for doing this is described in detail
in Section 4.6. All quantity manufacturers used
equipment basically the same as the laboratory
setup described in that chapter. Some of them
experimented with different types of trans-
ducers in order to get away from the limita-
tions of the displacement-type crystal pickups
used in the first design, which proved unsatis-

factory in service. Aside from erratic behavior,
the pickup acted as a microphone and the out-
put arising from its operation as such some-
times interfered with the voltage generated in
the pickup by the vibration under investiga-
tion. At least one manufacturer used a dynamic
pickup in conjunction with a simple RC net-
work to convert the output, normally propor-
tional to velocity, to a value proportional to
displacement. This same manufacturer also ex-
perimented with a very rigid (high resonant
frequency) nose mount in the balancing fixture
instead of the low-period flexible mount de-
scribed in Chapter 7.

After the approximate amount of the unbal-
ance had been determined, together with the
angular relationship of the heavy point to a
fixed mark on the vane, weight was removed by
either clipping the edges of the metal vane with
a pair of tin snips or drilling small holes in the
appropriate place on the plastic vane. Because
of the greater number of discrete points at a
maximum radius from which weight could be
removed on the metal vane, these were very
much easier to balance in production than the
plastic ones.

65 POWER SUPPLY AND ARMING

For the purpose of discussion, the power
supply and arming systems used on radio-type
proximity fuzes can be divided into two general
classifications. In one type of fuze, the power
supply and arming system was combined in a
separate subassembly adaptable to being farmed
out to subcontractors and later assembled to the
fuze head at the plant of the principal manu-
facturer. An example of a power supply of this
character is shown in Figure 22. In the other
general classification, the components making
up the power supply are distributed throughout
the fuze assembly in such a manner as to pre-
clude the identity of the power supply as a sepa-
rate assembly. Examples of this type of con-
struction are shown in Figures 23 and 24. In
Figure 23 the generator proper and its turbine
drive was installed below the fuze head, while
the rectifier and filter condensers associated
with it were distributed in various portions of

SECRET

POWER SUPPLY AND ARMING

267

the upper cavity containing the other electronic
components. In Figure 24, the generator, with
its driving turbine, is contained in the nose and

Figure 22. Integral power supply (left) as re-
ceived from outside manufacturer for assembly
into radio proximity fuzes. Oscillator-amplifier
assembly is shown at right.

the rectifier and filter components in the main
body of the fuze.

651 Requirements

The performance desired of a power supply
can be easily specified in terms familiar to the
electrical industry. No unfamiliar concepts are
involved. The supply must deliver plate, fila-
ment, and bias voltages which fall within speci-

Figure 23. Power supply and arming system
for T-82 fuzes.

fled limits over the expected range of speed
variation. The degree of filtering can be speci-
fied in terms of permissible modulation of the
plate supply.

Likewise, the mechanical specifications are
easily understood. The bearings must be ca-
pable of standing high-speed operation and the
arming system must perform its function with-
in a given number of revolutions of the gen-
erator shaft.

Since the fuze has such a short operating life,
it is permissible to overload some of the electric
components. This is a particularly fortunate
circumstance because the limited space avail-
able does not permit the use of components hav-
ing the safety factors usually specified.

Figure 24. Mechanical parts and power supply
for T-132.

65,2 Procedures

Generator Construction

The housing used in early models of genera-
tors were of molded Bakelite. This material
proved to be unsatisfactory because of difficul-
ties in maintaining the desired dimensional
tolerances and was abandoned after unfavor-
able pilot production experience in favor of
either die cast or stamped and drawn frames.
A power supply using die cast housings is
shown in Figure 25 and one using a drawn case
is illustrated in Figure 26. The die cast genera-
tor housing required a relatively large amount
on machine work on the rough casting in order
to make it usable. The tooling designed for this
purpose was somewhat elaborate and ingen-
ious. In one manufacturer’s plant, four mul-
tiple spindle drilling heads were used, each one
equipped with a five-position indexing platform
with provisions for rapid positioning and lock-
ing of housings in position. The four heads per-

268

PRODUCTION

formed 60 operations on each housing and pro-
duced one completely machined generator
frame every 90 sec. On each piece there were
18 drilling, 27 counter-boring, 8 counter-sink-
ing, and 12 tapping operations.

POWER SUPPLY ASSEMBLY

the bearing cups, which are clearly shown in
the illustration.

Both sleeve and ball bearings were used in
the production model generator. The sleeve
bearings were of sintered porous bronze. In
order to make sure that an adequate supply of
lubrication was available, even after long stor-
age periods, some manufacturers used satu-
rated wicks in connection with these sleeve
bearings. Later, ball bearings were used when
the supply of such bearings became adequate to
support the heavy requirements of the fuze pro-
duction program. These bearings make possible
a “tighter” generator assembly, end play and
side play being reduced to a minimum. In order
to keep production up and costs down, most
manufacturers used sleeve-bearing fits some-

STATOR ROTOR REGULATION NETWORK

Figure 25. Power supply using die cast gener-
ator.

The drawn shell housing was probably the
most feasible from a production standpoint,
and the majority of power supplies built used
this type of construction. This type of genera-
tor is shown, in considerable detail, in Figure

Figure 27. Details of generator using drawn
shell construction.

POWER SUPPLY ASSEMBLY

GENERATOR

REGULATION NETWORK

STATOR ROTOR

* RECTIFIER FILTER

Figure 26. Power supply using stamped and
drawn shell for generator.

27. The shell consists of two mating drawn
pieces which contain within themselves all load
holes and stator spacing and locating surfaces.
The only machined piece used in the shell was

what looser than was generally considered de-
sirable. The use of ball bearings also provided
a larger margin of safety against bearing fail-
ure during production testing. Where ball bear-
ings were used, cup-shaped beryllium-copper
spring washers were used to take up end play
and provide for slight dimensional differences.
Since the amount of take-up varied within wide
limits from fuze to fuze, additional shimming
was provided by a series of punched Bakelite
washers approximately 0.010 in. thick.

Generator coil construction took two general
forms, one using six bobbins, illustrated in Fig-
ure 24, and the other a single serpentine coil
assembly containing both plate and filament
windings such as illustrated in Figures 26 and
27. Two serpentine windings were used on
some types of generators, the second winding
passing over the opposite side of the stator
pole. The cost of the bobbin-type winding was

POWER SUPPLY AND ARMING

269

greater than the single serpentine coil. In addi-
tion, the bobbin-type construction had several
other disadvantages. It was necessary to pro-
vide six molded bobbins, wind each bobbin with
two separate windings, and interconnect the six
in the proper manner. Compared to this, con-
struction of the serpentine winding was rela-
tively easy and inexpensive. The plate and fila-
ment windings were wound one on top of the
other in a simple collapsible wooden form.
After removal from the form, the coils were
taped on the same type of equipment used for
taping small motor windings. The taped wind-
ing was then shaped in a simple fixture
having interleaved castellated projections
which pressed the taped coil into the charac-
teristic serpentine shape. The coil was then
slightly distorted and inserted in the stator and
expanded into position. The entire stator stack
was then vacuum impregnated.

The impregnation of both bobbin and serpen-
tine-type stators was essentially the same. The
stator assemblies were placed on a rack and
dried in an oven at 250 F for approximately
one-half hour. While still hot, the rack was
immersed in a container of suitable varnish
(Irvington Varnish and Insulation Company
No. 9 Clear Drying Varnish). For proper pene-
tration, it was necessary to hold this varnish at
a specific gravity of 0.855, naphtha or benzene
being used as a thinner. The container with the
immersed coils was then placed in a vessel and
evacuated to at least 25-in. mercury vacuum
for 15 min. The vacuum was then released and
the stator assemblies removed from the var-
nish, placed in a centrifuge, and the excess var-
nish extracted. The stators were then allowed
to air-dry at room temperature. Both bobbin
and serpentine coils were random wound.

Generator shafts were made of stainless-
steel ground precision finished stock, with the
worm cut on a standard thread grinder. This
worm was cut in one pass with a floor-to-floor
time of approximately 8 sec. After cutting the
worm, it was found necessary to de-burr the
machined portion. No method of generating the
worm was devised to get around this time-
consuming hand operation. The shafts were
held in the rotor insert by a knurled portion of
the shaft. This knurling increased the diameter

approximately 0.002 in. and provided a push
fit of the shaft into the hole in the rotor insert.

The Alnico rotors used were made either by
casting or sintering the Alnico material. By far
the larger number of generators produced em-
ployed cast Alnico IV rotors. In early models
of the generators, the soft steel or brass insert
engaging the shaft was held in place in the cen-
tral hole of the rotor by cerromatrix alloy. This
procedure proved unsatisfactory for two rea-
sons. First, the alloy has a very low melting
point and sometimes loosened from the heat
generated in the bearings by long test runs. The
mechanical problem of centering the hole in
the insert with respect to the outside diameter
of the rotor was solved after some trouble by
holding the outside diameter of the rotor and
the inside diameter of the bushing in a con-
centric collet-type fixture while the cerro-
matrix was poured in the space between the
rotor and hub. This method of holding hub was
abandoned later in favor of a solid soft steel
insert cast in the center of the Alnico rotor.
The cast blanks were next ground so as to have
the two sides parallel and to the proper dimen-
sions. These blanks were then centerless
ground to the proper outside diameter, after
which they were placed in a collet-type chuck
and the shaft hole drilled and reamed to the
proper size.

Some trouble was experienced in the begin-
ning of production with inability of the rotors
to stand high rotational speed. The manufac-
turers of the rotors solved this problem so suc-
cessfully that rotor breakage from this cause
was practically unknown toward the end of
the production program.

It was at first thought that rotors could be
held so close to the proper dimensions by the
sintering method of manufacture that some of
the grinding and sizing operations could be
eliminated. This, however, proved not to be the
case.

The rotors were magnetized in several dif-
ferent ways. Practically all manufacturers used
a fixture having six retractable pole pieces
around each of which was wound the magnetiz-
ing coil connected in such a manner as to pro-
vide opposite magnetic polarity to adjacent
poles. Some manufacturers advanced and with-

270

PRODUCTION

drew the pole pieces with all cams actuated
simultaneously by one handle. Other manufac-
turers used a fixture in which each pole was
attached to the piston of a small air cylinder
with the pole pieces advancing with air pres-
sure and withdrawing through the action of a
spring built into the cylinder assembly. Figure
28 shows a fixture of this type. Some manufac-
turers magnetized rotors using a bank of stor-
age batteries as a high-amperage low-voltage
source. Considerable difficulty was had with the
electric contacts because of the large currents
they were required to pass. A more satisfactory
method of doing the job was worked out by
some manufacturers who charged a bank of con-
densers to approximately 300 to 400 v using a
small receiver-type power supply to furnish the
charging current. These condensers, having a
total capacity of several hundred microfarads,
were then discharged instantaneously through
the magnetizing coils. A grid-controlled gas-
eous discharge tube was used to trigger the dis-
charge and by its unilateral conduction prevent
oscillation. Another manufacturer used a fix-
ture employing somewhat the same idea as the
one just discussed but discharging the con-
densers through the primary of a step-down
transformer having a secondary of a very few
turns which was coupled to single-turn mag-
netizing coils made of heavy copper strap. All
these devices served to saturate the magnet
material in a satisfactory manner.

Arming System

The mechanical construction of the arming
system employed in various production fuzes
have been adequately covered in Chapter 4.
Since no particularly new procedures were in-
volved in the construction of these components,
no discussion of them is considered necessary in
this chapter.

Electric Components

The electric components in a power supply
consist of filter and regulating condensers, re-
sistors and the selenium rectifier assembly. The
resistors used were standard commercial items,
and as mentioned previously all were operated
under conditions where the rated dissipation
was exceeded. The resistor in the regulating

network normally rated at x/± watt was called
on to handle in some instances as much as 3
watts for the short operating time of the fuze.

The filter condensers used in a majority of
the fuzes were specially designed and some diffi-
culty was experienced at the beginning of the
production program in obtaining a satisfactory
product. The condensers were built around a
hollow tube through which the slow-speed shaft

Figure 28. Typical fixture for magnetizing
rotors in generator power supplies (Globe-
Union, Inc., photograph).

passed. Two types of construction were em-
ployed. One type employed a simple construc-
tion in which condenser sections manufactured
in the conventional manner were dropped in
place between two concentric cardboard tubes.
These sections were then interconnected and
the whole assembly potted. Since there were
voids between the condenser sections, this was
not the most effective way to utilize the space
available. Nevertheless, the first manufacturers
contacted felt that this design was more fea-
sible from a production standpoint than the
second type to be described. In this second type
of construction, later used by all manufac-
turers, the two filter sections and the regulat-
ing condenser were all wound in one operation,
the leads being brought out by means of tabs
laid between turns. While the working voltage
of the condensers was only 150 volts, the test
voltage was 300 volts. This necessitated a two-
paper construction. Manufacturers, however,
were able to build into the space available suffi-
cient capacity for the purpose.

"SECRET

POWER SUPPLY AND ARMING

271

In fuzes where the filter and regulating con-
densers were in the main body of the fuze in-
stead of a separate power supply, small sections
of conventional construction were employed.
One fuze used a hermetically sealed oil-filled
unit.

In the early developmental stages, it was pro-
posed to use copper oxide rectifiers primarily
because small buttons of a suitable size were
already available and no new techniques had
to be devised to produce a size suitable for the
fuze. Because of the unsatisfactory temperature
characteristics of copper oxide rectifiers, these
were soon discarded in favor of selenium rec-
tifiers. When first approached on the proposi-
tion of producing a rectifier suitable for fuze
applications, the manufacturer, who at that
time was the largest producer of such devices,
expressed considerable doubt as to whether a
selenium rectifier button could be produced in
the size required. Selenium rectifier elements
had never before been produced in anywhere
near the quantity under discussion.

At the time, all selenium rectifiers were made
with a central hole through the disk for a
mounting stud which held the stack in compres-
sion. Engineers at the Bureau of Standards
proposed a wholly novel type of construction
which was immediately adopted and placed in
quantity production. As can be seen from Fig-
ure 29, there is no center hole in the disk. The
active area of the disk is the center depressed
area. Contact to this active area is by means of
a low melting point metal coating sprayed so
that it extends up the side of the depression
and overlaps the top to a distance of approxi-
mately y16 of an inch. This overlapping area
contacts the base metal of the next disk. The
whole assembly is contained in a suitable holder
under compression from a small spring which
applies approximately 6-lb pressure.

The manufacturer first approached as a sup-
plier for these rectifier disks deserves consider-
able credit for the development of the manufac-
turing process and the clever tooling worked
out. The method of manufacture and tooling
was adopted with some modifications by the
other manufacturers engaged in the produc-
tion program.

In the production of the rectifier disks a

sheet of a base metal was first sand-blasted or
chemically treated to provide a surface to which
the selenium would adhere. These sheets, each
one large enough for approximately 100 recti-
fier buttons, were then punched out in such a
manner that the portion which eventually be-
came the button was projected halfway through
the metal. At the same time, register holes
were punched for aligning the plate in the
fixtures subsequently used. The plate was then
covered with a mask which left visible only the
upraised round portions and selenium powder
applied to the exposed surfaces. The plate was
then placed in an oven and heat-treated to
change the powdered selenium to a suitable
form. Various manufacturers used different
procedures for this step in the manufacture of
rectifiers. In most cases it was considered a
trade secret which they preferred not to dis-
cuss. After the above treatment, the plates were
placed in another fixture and a covering of
paper cemented to the tops of the buttons. In
the paper there was a hole which registered
exactly in the center of each button. Over this
another mask was placed which contained a
slightly larger hole. A low melting point alloy
similar to Wood’s metal was then sprayed over
the top of the masks. This metal formed the
conducting medium contacting the center of the
selenium button and extending up the sides
of the recess and overlapping the edges.

After spraying, the mask was removed and
the plate transferred to a fixture containing
a multiplicity of small plungers, each one con-
tacting the small area of counterelectrode ma-
terial for a disk. In series with each plunger
was a resistor serving to limit the current flow-
ing to the button during the electroforming
process. As the resistance of the button was
built up during the formation of the barrier
layer, the voltage drop across the button be-
came higher and higher until the desired re-
verse current resistance was attained. After
electroforming, the plate was placed in an accu-
rately registered die and all the buttons
punched out of the plate in the finished form
shown in Figure 29.

Another supplier of rectifiers used a method
of manufacture which resulted in a superior
end product, particularly with respect to uni-

SECRET

272

PRODUCTION

formity. Deposition of the selenium on the base
metal was accomplished by evaporation in a
high vacuum and subsequent curing at the
proper temperature to obtain the desired crys-
talline form. Plates of base metal large enough
for about 100 rectifier disks were processed in
the evaporation chamber. They were then cov-
ered with a mask of high-quality paper per-
forated with small holes to outline the actual
areas, the paper being held in place with a

rectifiers and tested as a complete assembly. In
the early days, considerable trouble was had
with defective buttons made by the first process
described. Figures 30 and 31 show some typical
defects. The illustrations are self-explanatory.

Contact to the buttons was effected by means
of small metal tabs or flags interleaved between
buttons and projecting through the sides of the
case. One manufacturer used small coiled wire
forms for this purpose in place of flags.

Figure 29. Rectifier assembly using selenium disks.

coating of thermosetting plastic. A counter-
electrode material similar to that previously
mentioned was sprayed over the entire area of
the disks by using a skeleton of a previous
punching operation as a mask. By extending
the counterelectrode material over the whole
surface of the paper, the contacts between
adjacent disks in the finished rectifiers were
maintained continuously during severe shock
and vibration. The problem of microphonics in
rectifiers was resolved by this expedient.

The buttons were assembled into completed

66 MISCELLANEOUS PRODUCTION
TECHNIQUES

As was to be expected, each manufacturer
used techniques which had proved most desir-
able in his experience in the manufacture of
other electronic equipment (radio receivers in
most cases). Some manufacturers presented
the smaller assemblies to a fixed soldering iron
while others kept the units in a holding fixture
and applied the soldering iron to the work.
Each method has its own advantages and dis-

MISCELLANEOUS PRODUCTION TECHNIQUES

273

advantages and just which is best depends on
the nature of the operation.

Considerable difficulty was had in obtaining
solder having high tin content due to the scar-
city of this metal. As a result, some manufac-
turers were forced to use low tin alloys which
made the soldering operation somewhat more
difficult. Actually, a solder having 63 per cent
tin and 37 per cent lead has the lowest melting

tin introduces problems in obtaining properly
soldered joints.

Most solder specifications are written to
allow a ±5 per cent variation in the percentage
of tin used. Since tin was not only expensive
but scarce during World War II, most of the
40-60 solders used actually had less than 40 per
cent tin content. This necessitated the use of
more heat on soldered joints, with the added

GOOD CELL PARTIAL RING

OFF CENTER BURNED
RING

THIN

RING

WASHER

DEFECT

Figure 30. Typical defects in selenium rectifier disks.

point of any lead-tin alloy, 183 C. The plastic
range, i.e., the range of temperature in which
the solder is in molten form, is also shortest
with this alloy as is to be expected. The follow-
ing tabulation shows the melting point and
plastic range of various solder alloys.

Tin-Lead

Melting point
(degrees
centigrade)

Plastic range
(degrees
centigrade)

40-60

265

45-55

252

50-50

239

55-45

223

60-40

202

63-37

183

70-30

195

In normal times, most manufacturers prefer
to use a solder having at least 50 per cent tin
and some insist on a 60 per cent tin alloy. The
necessity of using alloys of 35 and 40 per cent

bad effects on resistors and condensers. There
was also considerable danger that the leads
might be displaced while the solder was taking
such a long time to reach a solid state. The use
of high tin content solders is particularly de-
sirable when soldering to metal parts embedded
in thermoplastic materials.

Another particularly troublesome point was
the poorly tinned lead wires on the resistors
of some manufacturers. Resistors were often
received with a waxy gum on the leads that
made soldering to them particularly difficult.
No really satisfactory method of cleaning this
material from the resistors was evolved.

It is interesting to note the various methods
used by different manufacturers in handling
fuzes along a production line, particularly after
the oscillator and amplifier assemblies had been
combined. Some manufacturers placed the fuze

274

PRODUCTION

MANUFACTURER I MANUFACTURER I MANUFACTURER
A I B I C

UNVARNISHED I VARNISHED

GOOD CELL | v ]
INSUFFICIENT ALLOY |
EXCESSIVE ALLOY | / • ;

BURNED | ■ [

OFF CENTER| J

DEFECTIVE WASHER I
TOO LARGE 1 :^f|gl§|

HIGH RIM | *

NO WASHER OR ALLOY if

ASSEMBLY WITH BEADS

CELLS FROM
ASSEMBLIES
SHOWING BEADS

Figure 31. Typical defects in selenium rectifier disks from three different manufacturers.

PRODUCTION TESTING

275

in a simple wooden fixture which was passed on
by hand to the next operator. Others used a
trough in which the whole fixture was a sliding
fit. The operator would remove the fixture
from the trough, perform the necessary opera-
tion, replace the fixture in the trough, and
shove it on to the next operator. A portion of

Figure 32. Assembly line for oscillator units.
Oscillators are moved along trough shown on left
side of photograph (Emerson photograph).

an oscillator assembly line is shown in Figure 32
with the trough used to pass on assemblies
shown at the left. Another used a conveyor belt
slowly moving along in front of each position.
To this conveyor belt was fastened a fixture
holding the unit. The operators were required
to perform the operation while the units were
on the move, so to speak. Overhead conveyors
were also used. Figure 33 shows such a system
feeding finished units to a final test area in the
plant of one manufacturer. Figure 34 gives a
close-up of a final test position showing the
small pockets attached to the belt in which the
fuzes were held.

6 7 PRODUCTION TESTING

The design of test equipment for proximity
fuzes is covered in Chapter 7 of this report.
Test equipment development for the fuze pro-
gram was the responsibility of the Central
Laboratory of Division 4 at the National
Bureau of Standards [NBS]. Since the speci-
fications for the fuze were written around tests

performed on equipment of NBS design, most
manufacturers followed the NBS designs in
the construction of production test equipment.

More testing was done in pilot production
than was deemed necessary or desirable in
quantity production. Not only were more tests
conducted, but it was necessary to record in
considerable detail data on the performance of
every unit in order that the known characteris-
tics of fuzes might be correlated with subse-
quent performance of the unit in field tests.
However, when meters have to be read to an
exact value and perhaps recorded, the process
takes more time than would be feasible in mass
production. For production purposes, prac-
tically all indicating instruments can be marked
with go and no-go limits and fuzes tested in
a very short time.

In most cases manufacturers followed a test
schedule similar to the following. Oscillator
assemblies after completion were tested for
(1) carrier frequency, (2) diode voltage (in
the case of oscillator-diode [OD] units), and
(3) grid voltage (in the case of reaction grid

Figure 33. Assembly line for radio proximity
fuzes showing overhead conveyor for moving com-
pleted fuzes to final test position (Emerson
photograph) .

detector [RGD] units). Amplifier assemblies
were given a preliminary test after construc-
tion, principally to see if the circuit was func-
tioning. After interconnecting the oscillator
and amplifier assemblies and before potting, a
rather complete check was made on the com-
bined “head,” the following information- being

EGRET

276

PRODUCTION

noted on each unit: (1) millivolts (input to
amplifier) to fire (the thyratron) at the peak
audio frequency, (2) millivolts to fire at two
frequencies spaced from the design center fre-
quency in such a way as to serve as an indica-
tion of the shaping of the amplifier, (3) peak
audio frequency, (4) oscillator frequency, and
(5) diode or grid voltage.

After potting, most manufacturers tested the
fuze head to determine whether or not any sig-

loads. They were also checked to observe volt-
age regulation (of the power supply) over a
specified speed range. Most manufacturers used
an oscilloscope connected across the high-
voltage output which served in some instances
as a visual indication of erratic behavior which
would not otherwise have been detected.

In the minds of some manufacturers was a
well-defined suspicion that power supplies were
a source of noise and several manufacturers

Figure 34. Final test position on production line. Fuzes are shown arriving at position via overhead
conveyor belt (Emerson photograph).

nificant changes had taken place because of
potting. In some cases, this test was abandoned
after experience had shown that the number of
units rejected at this test position was negli-
gible.

Power supplies made by outside suppliers
were given an incoming inspection at the plant
of the principal contractor, even though the
unit had been checked as satisfactory by the
original manufacturer. Units were checked for
A voltage, B voltage, and C voltage at rated

had under way at the end of the production
program the design of equipment intended to
segregate these noisy power supplies. Most of
these devices took the form of a transient de-
tector built around conventional thyratron cir-
cuits, a noisy unit being indicated by either a
visual or audible signal.

The final acceptance test given completed
fuzes was most complete. The acceptance or
rejection of the unit was based on measure-
ments of the following values: (1) carrier fre-

PRODUCTION ACHIEVEMENT

277

quency, (2) diode voltage or grid voltage, de-
pending on the type of unit, (3) millivolts to
fire at peak audio frequency, (4) peak audio
frequency, (5) A voltage, (6) B voltage, (7) C
voltage, and (8) effective critical voltage of
the thyratron.

The last criterion (effective critical voltage)
was determined by establishing a fixed bias on
the thyratron grid by means of the special cir-
cuits described in detail in Chapter 7 and run-
ning the unit over a specified range of speeds.
If the unit fired during this run, the bias volt-
age was raised a fixed increment and the speed
run repeated. Units which required a holding
voltage greater than a specified value to prevent
firing over the established speed range were
rejected for noise.

Most manufacturers maintained a rework
department staffed by technicians familiar
with the operation and circuits of the fuzes and
various subassemblies. Subassemblies or com-
pleted units rejected at the various test posi-
tions were shunted to this rework department,
diagnosed, and repaired if the repair job was
deemed to be economically feasible.

The apparatus used at the various test posi-
tions varied in slight details from manufac-
turer to manufacturer, although basically the
circuit arrangements were the same. Some
manufacturers went the limit in designing in-
genious holding and connecting fixtures to ex-
pedite testing. Much use was made of air
clamping devices and multiple contact fixtures
wherein all connections were made to a fuze or

power supply simply by depressing one lever.
It is a well-known fact that no tool and fixture
designer likes to copy completely the design
used in another plant, and as a consequence the
variations in methods of accomplishing the
same end result were very interesting to ob-
serve. Since the design of holding fixtures and
the like for fuze production presents no prob-
lems that have not been met in the manufacture
of other electronic equipment, a complete de-
scription of the devices does not seem to be in
order.

PRODUCTION ACHIEVEMENT

The following information, taken from re-
ports from some of the major manufacturers
involved, shows the magnitude of the production
achieved.

Manu-

facturer*

Total fuzes
produced
315,000
247,138
255,996

Number of
Rate per month production

at peak of
operations
52,800
40,418
39,600

employees

involved

1,050

883

1,000-1,800

* Manufacturer A made the complete fuze in the plant, including
the power supply. Manufacturers B and C bought power supplies
from outside sources.

Figures are available from only one manu-
facturer of power supply assemblies. They
show a total of 490,150 power supplies made
with production reaching a peak of 60,000 per
month with 350 production employees.

SECRET

Chapter 7

LABORATORY TESTING OF FUZES*

71 INTRODUCTION

For the purpose of expediting design engi-
neering and production control, laboratory
tests were required to obtain pertinent per-
formance data. These tests and associated
equipment are described in detail in this chap-
ter.

The general outline of preceding chapters in
which the principal performance characteris-
tics and production problems were discussed
will be followed in this chapter. A description
of tests on the radio- and audio-frequency sec-
tions will be followed by a discussion of tests
on components and other relevant tests. In
addition, a brief outline of the tests used in a
typical quality control laboratory, and an out-
line of tests used in a typical pilot line are in-
cluded. As will be noted, the quality control
test line is in general the reverse of the pilot
line. This is obvious in that a quality control
laboratory receives a completely assembled
fuze, while a production line starts with com-
ponents and ends up with a completely assem-
bled fuze.

The emphasis of this chapter will be on the
general principles involved while testing, omit-
ting the theoretical discussion, since this is
covered in Chapters 2 and 3. It should be
pointed out that Chapters 2, 3, and 4 also in-
clude discussion of tests not mentioned here
because such tests were related to development
problems rather than the testing of finished
fuzes.

The preferred laboratory testing procedure
was to evaluate the performance of the sepa-
rate sections of the fuze, i.e., r-f, audio, detona-
tor circuit, and power supply, rather than to
attempt to devise an overall performance test.
The reasons for this approach were as follows :
Fuze failure can occur from inferior or sub-
standard performance from any of the various
sections of the fuze. Testing each section sepa-

a This chapter was prepared by Thomas C. Bagg and
Paul J. Martin of the Ordnance Development Division
of the National Bureau of Standards.

rately for conformance to requirements insured
reasonably good performance for the complete
fuze. If overall tests were used, inferior per-
formance of one section might be compensated,
and hence masked by extra sensitive perform-
ance of another section. For example, a fuze
which has an insensitive r-f section and a high-
gain amplifier will fire the thyratron with a
normal signal, since one section compensates
for the other. If a section is unusually sensitive,
the fuze may tend to become unstable and hence
the probability for malfunctions of the fuze is
greatly increased by a part which is out of tol-
erance.

Numerous attempts were made to devise an
overall test but none of them appeared to offer
the same assurance that fuzes would perform
as reliably in the field as when the individual
sections of the fuze were tested.

In designing test equipment, there were cer-
tain practical considerations to be taken into
account. From the standpoint of production,
equipment had to be designed to provide a
maximum of economy in time, effort, and ma-
terials, yet give the required accuracy and ease
of operation. The length of time the unit or sub-
assembly was under test had to be as short as
possible to conserve the life of component
parts, such as tubes, bearings, and gear trains.
Accuracy of the meters and other indicators
of the test equipment had to be kept as high
as possible by frequent calibration against suit-
able standards and adequate compensation for
humidity and temperature variations.

7 2 TESTS ON THE R-F SECTION
7,2,1 Measurements Required

There are three kinds of r-f assemblies to be
tested: oscillator diode [OD], reaction grid de-
tector [RGD], and power oscillating detector
[POD]. The parameters which determine the
performance of these assemblies are diode volt-
age (for oscillator-diode units only), oscillator

SECRET

278

TESTS ON THE R-F SECTION

279

grid voltage, plate current, and carrier fre-
quency.

As shown in Chapter 3, these parameters
vary because of variations in radiation resist-
ance upon approach to the target. However,
certain of these parameter variations are more

tions of loading and supply voltages, since in-
stability will produce a malfunction.

The preceding statements apply to measure-
ments on the r-f subassembly. When measure-
ments were made on the completed fuzes, it was
necessary to have the r-f section operate under
proper conditions of loading. The methods by
which these conditions were obtained in the
final test position are also discussed here.

Loading Requirements

The load presented to a fuze is composed of
resistive and reactive components which are
dependent upon the dimensions of the missile
and the frequency of the oscillator. The radio
frequency used is that which will give the re-
quired sensitivity and stability of the fuze for
the missile or missiles on which it is to be used.

Figure 1. Reference vehicles for testing prox-
imity fuzes. These represent, from left to right,
M-30 bomb, M-64 bomb, and 5-in. AR rocket.

significant for each fuze type, that is, diode
voltage for diode detectors, oscillator grid volt-
age for reaction grid detectors, and plate cur-
rent for power oscillating detectors. It should
be remembered, however, that all these param-
eters and carrier frequency are interrelated.
It is therefore necessary to measure not only
the steady-state values of these parameters but
also the rate of change with load of the signifi-
cant parameter for each fuze type. Such a
measurement is an indication of the r-f sensi-
tivity. Further, this section must be checked for
stability when operating under extreme condi-

Figure 2. Final test chamber for ring-type
fuzes.

Since it is inconvenient to measure the param-
eters which determine performance on actual
missiles in free space, some form of laboratory
test equipment had to be designed which would
give accurate values of these parameters under
simulated operating conditions.

To insure proper operation, free-space load-
ing conditions were used as the basis, or refer-
ence point, for all laboratory measurements.
As pointed out in Chapters 2 and 3, there was
an optimum frequency for each missile which
would give the required sensitivity, but, since

SECRET

280

LABORATORY TESTING OF FUZES

most of the fuzes had to operate on more than
one missile, a compromise on frequency was
made and a typical, or reference, missile chosen
for test purposes. The following table gives
the reference missile chosen for each fuze
type.15, 16, 39

Class of
projectile
Bomb
Bomb

Aircraft

rocket

Aircraft

rocket

Mortar

Frequency, fuze type

Brown frequency, ring-type
White frequency, ring-type and
all bar-type fuzes
Brown frequency, ring-type
and miniature rocket fuzes
T-5 and T-6

All-frequency mortar fuzes

Reference

missile

M-30

M-64

5-in. AR

M-9

M-43C

For convenience in calibrating laboratory
equipment, mockups of the missile which could

load because under light loading unstable units
were more readily detected.

In OD units, the reactive component of the
load, however, had to duplicate that of free
space, since any reactive load across the an-
tenna not only changed the value of the param-
eters14 but changed the operating point of the
oscillator (or diode circuit) in such a manner
as to reduce the r-f sensitivity. For example,
1 ^f of additional capacity across the nose cap
of a diode detector-type fuze reduced the volt-
age by about 8 per cent, resulting in a reduc-
tion in sensitivity of approximately 16 per cent.
The reactive component of the load was not
critical in RGD units (see Sections 3.1.1 and
3.1.2).

Figure 3. Compensated loading resistor on ring-
type fuze.

be easily suspended in free space were made
containing batteries and meters (see Figure 1).

To simplify testing further and to facilitate
correlation of the equipment in the various
laboratories, a load resistance was chosen to
represent approximately the free-space load
(see Section 2.7). 3,20 The value chosen repre-
sented a slightly lighter load than the free-space

7 2 3 Shielding

In order to prevent interaction between fuzes
or the influence of nearby objects in the radia-
tion field, it was necessary to shield the fuze
during tests. The type of shield used for T-5
testing was a 16-in. plate placed behind the
fuze in such a manner as to present the same
capacity as the missile.48 The important feature
of this type of shield was that it unloaded the
oscillator without detuning the diode circuit,
but, on the other hand, it was not an infinite
plane and r-f voltages were induced in adjacent
test apparatus.

Completely enclosing shields were used for
testing generator-powered fuzes. The excess
capacitance loading produced by the shield was
neutralized by inductive compensation.24 For
tests on the r-f subassemblies of these fuzes, the
shields were usually 2-ft cubical metal-lined
boxes. For tests on the completed ring-type
fuze, where tuning and sensitivity measure-
ments were not made, it was found more con-
venient to use a small but very heavy all-metal
chamber (see Figure 2).

72 4 Loading Devices

Since the shield unloaded the oscillator, a re-
sistive load was necessary to secure proper
operating data. The dummy load developed for

TESTS ON THE R-F SECTION

281

T-5 tests consisted of an Aquadag (colloidal
suspension of carbon) line drawn on Scotch
tape and placed across the antenna insula-
tor.5*48 A similar device was used for loading
bar-type fuzes. This was Uskon cloth, a com-
mercial product of 377 ohms per square, which
was satisfactory when properly located in the
2-ft shield box.51

It was found possible to obtain the required
resistance and reactance loading by the use of
a resistance-capacitance-inductance parallel net-
work loosely coupled to the fuze.49 In some in-

quency ceramic resistor upon which was wound
a coil whose distributed inductance and capaci-
tance was sufficient to tune out the unwanted
portion of capacitance introduced by the shield
and resistor. Such a resistor is illustrated in
Figure 3. A modification of this type of load
was used by one manufacturer when they used
resistance wire to wind the inductance.

These compensated resistors provided the
proper compensation throughout the frequency
band used, since their reactance variations with
frequency followed those of most fuzes and

Figure 4. Inductively tuned load for OD ring-type fuzes.

stances, a diode rectifier and tuning indicator
were included.13 When this network was tuned
to resonance, it furnished only the resistive
component of the load,20 while the reactive com-
ponent of the load was adjusted by the cou-
pling. The position or coupling of the loading
device relative to the fuze was determined by
trial and error to duplicate free-space loading
conditions.

For use in the 2-ft shield box, an inductance
was wound on an ultra-high-frequency resistor
to compensate for capacity excesses introduced
by the box and resistor. Such a compensated
resistor24 consisted of an IRC ultra-high-fre-

bombs. The usable frequency range of the com-
pensated resistor for T-30 fuzes was very nar-
row. Because the rocket on which T-30’s were
used was long and thin compared to bombs, its
free-space reactance variation with frequency
was in the opposite direction.39

In the final test position, measurements were
made of the overall stability of the fuze to ran-
dom noise. It was particularly important that
the loading introduce no errors in the measure-
ments. Errors could be introduced in two ways :
(1) vibration of the load caused by high-speed
rotation of the generator, and (2) increased
FM noise in the oscillator due to a low LC ratio

282

LABORATORY TESTING OF FUZES

in the inductive load.17-45 In order to increase
the LC ratio of the load (already low due to the
presence of the enclosure), tuning was accom-
plished by a variable shunt inductance rather
than with an added variable capacitance and
fixed inductance. Coupling to the load in the
test chamber was first made through a ring
which fitted around the antenna of the fuze.
This method of coupling was replaced by a disk
in front of the fuze in order to reduce the ca-

Figure 5. Disk load for bar-type fuzes. Central
tube and nozzle carries compressed air to drive
windmill.

pacity of the load and to reduce inductive
coupling between the oscillator and loading
coils21*43 (see Figure 4).

For RGD units, no variable tuning was nec-
essary, so that the means of loading simply be-
came a compensated resistor connected between
the coupling disk and the chamber.

The test fixture for bar-type fuzes was a
2-ft shield box with the fuze mounted on a
standoff to reduce capacity loading across the
dipoles. The length of the standoff was impor-
tant in that it became a resonant line at par-

ticular frequencies, lengths, and diameters.25* 36
In the 2-ft box, a 3y2-in. pipe 7 in. long gave
no spurious effects. The load consisted of either
a sheet of Uskon cloth alongside the dipoles
(Figure 37 of this chapter) or a disk of Uskon
cloth in front of the dipoles (Figure 5). The
disk did not require orientation of the dipoles.
To prevent contamination of the load by oil,
sludge, etc., the cloth was covered by a thin
sheet of Lucite or celluloid. A test chamber for
laboratory testing of complete fuzes with trans-
verse antennas was designed for use as a noise
reference fixture to evaluate the production test
boxes (Figure 6). This chamber used a tuned
resistance load similar to that described above.43

Some difficulty was experienced in correlat-
ing different test positions because of changes
in r-f resistance with frequency. This variation
was greatest with high-value resistors.55 To
overcome this difficulty a set of resistors was
arbitrarily selected as reference standards.

For units which had normal radiation resist-
ances of 100,000 ohms or greater, it was found
desirable and convenient to use no resistive
component other than that in the tuning ele-

Figure 6. Final test chamber for bar-type fuzes.

ment of the load. This type of loading was used
with T-51, T-82, T-132, and T-171 fuzes.

Several experimental test fixtures were con-
structed which used a quarter-wave line to ob-
tain the r-f load. Westinghouse (Baltimore)
used such a device for their T-5 fuzes. Philco
also used it for their T-50 production. Some

SECRET

TESTS ON THE R-F SECTION

283

work was done at the National Bureau of Stand-
ards on this type of loading, but not with par-
ticularly satisfying results except in the case of
a device for the measurement of absolute sensi-

Figure 7. Resistor for determining sensitivity
of bar-type fuzes.

tivity of an end-fed axially excited fuze (see
Section 2.12).

Loads based on parallel transmission lines
were also experimented with but not used to
any great extent.

tivity obtained from the formulas [equations
(4) and (6), Chapter 3]

d In R’

s-E-(r.-‘ }

Figures 3 and 7 illustrate loading resistors
used in sensitivity determinations. For certain
units where the normal load was in the linear
portion of the loading curve (see Figure 7,
Chapter 3), only two points, or resistors, were
necessary to determine the sensitivity. For fuzes
which had very high load resistance, uncom-
pensated resistors of 100,000 ohms and infinity
were used where [equation (16), Chapter 3]

s = Vm-V.

It was first thought unimportant to induc-
tively compensate sensitivity resistors for RGD
units,44 but it was found desirable to do so in
order to prevent shifting of the operating point
of the oscillator.

Sensitivity Test

The most direct method of measuring r-f
sensitivity is a pole test as described in Section
2.12.

This test was the only practical means of de-
termining the sensitivity of very early fuzes,
because the close coupling between the oscilla-
tor and diode circuits would throw the oscilla-
tor into unstable operation when the fuze was
unloaded to obtain Vm for the sensitivity for-
mula [equation (4), Chapter 3].

However, for production testing, this was
impractical and was used only as the stand-
ard for free-space loading and absolute sensi-
tivity measurements.

The use of compensated resistors in the 2-ft
shield box gave rapid and sufficiently accurate
data for sensitivity determinations. For such
determinations, the diode voltage, grid voltage,
or plate current was plotted against the natural
logarithm of the load resistance and the sensi-

726 Stability Test

In order to test fuzes for r-f stability, the
first test used for T-5 fuzes5 consisted of de-
creasing the r-f loading and tuning the unit
through resonance. If no discontinuities were
observed in diode voltage, plate current, or
carrier frequency, the fuze was considered
stable. Other indications of instability were
self -blocking (squegging)5 which was noted by
the presence of discrete side bands. A more
satisfactory stability test was devised where
an alternating plate voltage was applied to the
oscillator, which caused it to go in and out of
oscillation. The plate voltage was applied to
the horizontal plates of an oscilloscope, while
the oscillator grid voltage was applied to the
vertical plates ; this showed the grid voltage as
a function of plate voltage and readily dis-
closed any tendencies toward instability.

SECRET

284

LABORATORY TESTING OF FUZES

7,2,7 Carrier Frequency

Carrier frequencies were measured by loosely
coupled absorption-type wavemeters or ultra-
high-frequency receivers. Care was required
when using superheterodyne receivers to insure
that the true frequency was read and not that
of the image or some other spurious responses.
To maintain accurate calibrations, harmonics
of a standard 5-mc oscillator were used. These
oscillators were periodically checked against
WWV, the standard frequency station of the
National Bureau of Standards.

AUDIO TESTS

connected or blocked, the impedance which the
amplifier saw when looking back into the oscil-
lator would be altered and normal signals such
as filament hum coming from the oscillator
would be distorted. Any indicating device con-
nected to the amplifier output, i.e., thyratron
grid, was such that it did not affect the ampli-
fier characteristics. A high-impedance cathode
follower was usually used for coupling test
instruments to the amplifier output. Any con-
nection to the thyratron plate had to be such
that proper voltages were applied, since the
thyratron critical voltage was a function of its
plate voltage. Also, any firing indicator circuits
had to have current limiters so as not to weaken
or destroy the thyratron.

7,3,1 Measurements Required

The function of the audio portion of the fuze
is to select the proper signal and amplify it suf-
ficiently to actuate the trigger circuit (thyra-
tron). Laboratory tests required were there-
fore those necessary to determine the gain-

7 32 Input Circuits

It was necessary to use various types of input
circuits to meet the needs of different units.
With oscillator-diode units, the diode was
blocked while making amplifier measurements
to eliminate noise developed in the oscillator.1

|:| ISOLATION TRANSFORMER

AUDIO

OSCILLATOR

l:l ISOLATION
TRANSFORMER

AUDIO

OSCILLATOR

Figure 8. Schematic of audio input test circuits (OD, RGD, POD).

frequency characteristic as well as the peak
gain. Gain, as such, was not measured. Instead,
the input signal to the amplifier (in rms milli-
volts) required to fire the thyratron was used
to indicate amplifier quality. Every effort was
made to insure that test conditions were the
same as those which existed when the fuze was
in operation. Hence, if the oscillator were dis-

This was easily done by applying a negative
voltage to the diode and amplifier test lead. This
voltage had to be greater than the peak r-f
voltage developed in the diode circuit to com-
pensate for the rise in r-f voltage occurring
when the diode was blocked and made non-
conducting. Blocking the diode changed the
impedance which the amplifier saw when look-

AUDIO TESTS

285

in g back toward the oscillator.41 To correct for
this change of impedance, a resistor of 170,000
ohms was used in series with the blocking bat-
tery46 and source of audio voltage. This audio
voltage was obtained from a commercial audio
oscillator through a voltage divider in order
to permit metering of the voltage by rectifier-
type voltmeters. Typical audio input circuits
are shown in Figure 8.

Figure 9. Schematic of cathode follower and
d-c VTVM (used on amplifier output) .

For testing fuzes which contained no diode,
the r-f signals were eliminated by either dis-
connecting the oscillator from the amplifier
(which in some instances was inconvenient) or
using a low-impedance by-pass between the
oscillator and high-input impedance amplifier.19
In the case of choke51 or transformer input, it
was necessary to use the first method and ob-
tain the audio voltage through a circuit equiva-
lent to the fuze oscillator.

As pointed out previously, the filament hum
present in the oscillator output due to an un-
balanced filament supply had to be duplicated
by a voltage injected with the test signal. This
was done by raising the voltage divider above
ground potential by the amount of hum voltage
necessary to duplicate that present in the oscil-
lator output. The phase of the hum so injected
had to be within 10 degrees of that in the oscil-

lator output. Care to maintain the normal value
of filament hum appearing at the thyratron
grid was also necessary in order to obtain
proper effective critical voltage for the thyra-
tron (see Section 3.3.5). 10’ 47

7 3 3 Output Circuits

Since the amplifier load consisted of an RC
network which controlled the high-frequency
cutoff and phase of the feedback voltage, it was

Figure 10. Cathode follower impedance chart.
Measurement of cathode follower input imped-
ance by use of accurate 30- and 90-megohm re-
sistors.

Apply a convenient voltage (10 volts, 200 c) directly to
cathode follower input lead and observe output E\ on
vacuum-tube voltmeter. Apply same voltage through series
resistor of 30 megohms and note reading Eo ( 30 ) . Repeat
foregoing step with 90-megohm resistor and note reading
(90). Form the ratios

Eq (30) and Eo (90)

E i E\

Locate point on graph corresponding to two ratios and find
impedance by interpolating between curves of constant
impedance.

essential that any test instrument connected to
the amplifier output would in no way alter the
amplifier output impedance. For this purpose,
cathode followers were used for coupling to
commercial voltage indicators, such as volt-
meters, oscilloscopes, or magic-eye tubes. The

SECRET

286

LABORATORY TESTING OF FUZES

impedance of the cathode followers used was of
the order of 50 megohms at 200 c and 12
megohms at 1,000 c. Values in excess of 40
megohms at 200 c and 10 megohms at 1,000 c
caused negligible errors in the measurements
and these values were introduced as specifica-
tion limits for acceptance testing.

Impedance losses in test leads between the
amplifier and cathode follower due to capacity
to ground frequently caused difficulty in meet-
ing the specification limits. By using a shielded
lead where the shield is connected to the cath-
ode of the follower tube, this parallel impedance
loss can be greatly reduced (see Figure 9).
Figure 10 illustrates a method of rapidly de-
termining the input impedance of the follower
when the capacitive components of the lead and
tube are considered.

Thyratron Tests

Effective Critical Voltage. The highest nega-
tive grid biasing voltage which will fire the
thyratron is called the critical voltage. (The
critical voltage is, of course, different for a-c
operation of the thyratron filament than for
d-c.) The term normal critical voltage was ap-
plied to the highest negative biasing voltage
which would fire the thyratron in the operating
fuze assembly with microphonic noise from the
oscillator blocked. The usual procedure for
measuring normal critical voltage was to block
the oscillator and inject (at the amplifier in-
put) a ripple signal equivalent in magnitude
and phase to that ripple from the oscillator
filament. The term effective critical voltage was
applied to the highest negative biasing voltage
which would fire the thyratron when the fuze
was completely operating, that is, when micro-
phonic noise from the oscillator was passed on
to the thyratron grid. In the generator-powered
fuzes, these measurements were made with the
generator running (driven by an air jet
directed on the vanes) so that microphonics
induced by the rotating system would show up
as a change in the effective bias of the thyra-
tron.

In making measurements of effective critical
voltage, it was not possible just to reduce the

applied bias on the thyratron, since this bias
was also applied (through a voltage divider) to
the pentode.11 Such procedure would have
caused changes in amplification resulting in
other-than-normal hum voltage at the thyra-
tron grid. A high-impedance positive voltage
source was therefore applied directly to the
thyratron grid along with a high-impedance
voltmeter. It was also necessary in all of the
equipment to insure that there was very little
coupling between input and output test circuits,
since any such coupling would appear as paral-
leling the gain-control condenser (C8 in Figure

THY PLATE IOPOO OHMS
© VVAA

0.1 MF

+135 V

THY PLATE 10,000 OHMS

r — dr-i

i-iY

t^rr

4 WATT
NEON

4=0-1 <> |.0

ADD FOR TESTING
RC ARMING CIRCUITS

Figure 11. Firing indicator for thyratrons:
for tube and unit testing, except units with RC
arming (top) ; brighter flash, satisfactory only
when a-c signal is used on thyratron grid
(bottom) .

26 of Chapter 3) and seriously affect gain ad-
justment or amplifier performance.

A type test (later a production test) was re-
quired to determine if the actual bias at the
thyratron grid was approximately equal to the
applied C bias. Leakages through the output
coupling condenser to the plate supply, through
the potting compound to ground, and along the
component surfaces, tended to reduce the bias
at the thyratron grid. Such leakages, if present,
caused unstable and erratic performance.

Firing Indicator. The most convenient firing
indicator consisted of a neon bulb and RC net-
work where the neon fired either on discharge

SECRET

STABILITY TESTS

287

or charge of a condenser (Figure 11). Such a
circuit was advantageous in that it was simple,
unharmful to the thyratron, and quenched so
quickly that the thyratron would recover for
rapid testing. Other types of firing indicators
were developed and used, ranging from a simple
lock-in circuit which remained operative once
the thyratron fired, to circuits which rang bells,
flashed lights, etc.

Routine tests for passing surge currents
through the thyratron were frequently dis-
cussed but seldom used outside the laboratory
since the failure of a thyratron in a fuze to pass
ample current was rarely reported.

74 STABILITY TESTS

7,41 Purpose of Stability Tests

A major cause of malfunctioning of fuzes in
field and service tests was noise or micro-
phonics in the electronic assemblies. Accord-
ingly, considerable effort was made to devise
laboratory tests which would show up or sort
out noisy units. Since noise was usually pro-
duced by the vibration of loose or weak parts,
either in the circuit or in the tubes, the testing
methods employed shaking or shocking tech-
niques. The ability of a fuze to withstand vibra-
tion was considered as a measure of its sta-
bility.

Various methods were used to indicate the
stability of a fuze under vibration: (1) the
peak noise voltage at the thyratron grid, (2)
the highest negative bias applied to the thyra-
tron grid, which would cause firing, i.e., effec-
tive critical voltage under the selected condi-
tions of vibration, and (3) the difference be-
tween the effective critical voltage and the
thyratron bias voltage, i.e., noise margin.

7,42 Methods of Producing

Vibration or Shock

Laboratory methods of inducing vibration
in fuzes attempted to duplicate (in a crude

way) the vibrations experienced by the fuze on
a missile in flight. It is well known that air
turbulence and fin flutter produce intensive vi-
bration in missiles and these, of course, will be
transmitted to the fuze.

The first vibration or shock method employed
to select stable fuzes was the simple expedient
of striking the fuze (or rather a test missile in
which it was mounted) with a club. This
method was strictly qualitative, but field scores
were greatly improved by discarding fuzes
which showed excessive noise signals when hit
with the club.

The club test for fuze stability was refined by
producing the shock with a calibrated pendu-
lum and employing a standard mount for the
fuze.5 Although the shock test produced little
reliable quantitative data, it did lead to consid-
erable improvement in tube design and to im-
provements in the technique of anchoring com-
ponents in the r-f section. All T-5 fuzes were
subjected to the pendulum shock test.

With the advent of generator-powered fuzes,
an additional source of vibrational energy
appeared through slight unbalance of the high-
speed rotating system. Analysis showed that
such an unbalanced rotating system was the
best microphonics testing device, because it
produced exciting forces in all directions in a
plane, because the entire range of exciting fre-
quencies could be easily covered, and because
the exciting forces were maintained long
enough to build up peak amplitudes at resonant
frequencies.9’ 23 The weaknesses of the shock
test were the presence of directional effects and
the inability of a single shock to build up vibra-
tion amplitudes at resonant frequencies.

When a fuze was mounted in its adapter or
encasing can on a bomb, unbalance in the rotat-
ing system produced vibrations of large ampli-
tudes because the bottom of the adapter acted
as a resonant diaphragm in the frequency
range of 20,000 to 35,000 rpm (see Figure
12). This method of producing vibration was
used in the final test fixture where an adapter
was tightly screwed into a massive test cham-
ber. A shoulder of at least 0.010 in. was cut
on the bottom of the adapter to simulate mount-
ing on a bomb ogive (see Figure 12). To per-
mit uniform testing and to compensate for

1ECRET

288

LABORATORY TESTING OF FUZES

adapter differences, the bottoms of the adapters
were cut so that resonance occurred at approxi-
mately 25,000 rpm.

C Shaft shield

D Generator rotor (unbalanced)

E Gear train (binding)

F Adapter can bottom (resonant diaphragm)
G Ogive of bomb

A Adapter can bottom (resonant diaphragm)
B Shoulder greater 0.010 in.

C Test fixture

Figure 12. Resonant mountings: fuze on bomb
illustrating pertinent vibration elements (top) ;
test adapter (bottom).

Numerous difficulties were encountered in
calibrating such a mount. A study of adapters
showed that the mechanical properties of the

diaphragm were not uniform. This led to dif-
ferent amplitudes and changes in frequency
during use, which were probably caused by cold
working of the metal and fatigue. In addition,
an analysis of such a resonant mount in a test
chamber revealed numerous uncontrollable var-
iables.37 As a temporary expedient, an assembly
containing a rotating system of known unbal-
ance was used to calibrate test fixtures. Such
assemblies were called reference vibration
heads and are shown in Figure 13.

As propeller balancing techniques improved,

Figure 13. Reference vibration heads used to

calibrate vibration response of test fixtures.

self-excited resonant vibration systems lost
their ability to test adequately stability and
other methods were sought. One system pro-
posed was the addition of a known unbalance to
the propeller and the use of a soft mount to
overcome the difficulties inherent in a resonant
system. This system appeared to have a number
of advantages for obtaining a qualitative meas-
ure of unit stability.37 Development of this
method was not completed at the end of World
War II.

Many types of external vibrators (in con-
trast to the internal source of vibration in the
fuze's rotating system) were designed and
tried. The first, the rotary vibrator (Figure
14), followed the principle mentioned above,
that is, an unbalanced mass was rotated at high
speeds.9 The vibrator housing was supported on

SECRET

STABILITY TESTS

289

a soft mount with the fuze to be tested on the
other end. Bearing failure, poor high-speed
motors, and the small mass which such a sys-
tem could vibrate, limited the use of the rotary
vibrator to components (particularly tubes),
head assemblies, and miniature fuzes.

Several models of a vibrator using a process-
ing ring operated with some success (Figure
15). One contractor turned the unbalanced tur-
bine 90 degrees, getting vibration in a vertical
plane. Models in which a ball was driven at
high speed around a race showed promise38
(Figure 16). Another device which proved un-
satisfactory consisted of four diaphragms 90
degrees apart around the unit mount, the dia-
phragms being actuated by a rotating valve on
a high-pressure air line. The impedance of the
air supply lines was so great that very little
energy was delivered to the unit, and the
scheme was also unsatisfactory at high frequen-
cies. None of these vibrators was developed for
production testing by the end of World War II.

Figure 14A. Rotary vibrator. Schematic section
of head assembly mounted on mechanically driven
vibrators.

7‘4'3 Other Noise Sources

In addition to noises caused by microphonics,
there were other sources of noise in the fuze
which had to be eliminated (see Section 3.1.5).
In the later part of T-5 production, extremely

Figure 14B. Miniature fuze mounted on air-
driven vibrator; operates on principles illustrated
in Figure 14A where air turbine is unbalanced.

290

LABORATORY TESTING OF FUZES

Figure 16. Ball race vibrator.

erties of the pentode, while the reliability of the
detonator circuit was no better than the de-
pendability of the thyratron. Because the tubes
were tiny, complex, and difficult to manufac-
ture, it was necessary to work out careful per-
formance tests in order to insure that the
tubes would be satisfactory.6* 35> 53 Here we will
confine the discussion primarily to a listing of
the properties which were measured.

As indicated in Section 7.4, one of the most

be met in turn by setting close tolerances on the
components. This section outlines the types of
tests which were used to select the most im-
portant components. Details of the tests are
generally not given ; instead, reference is made
to source material in the bibliography.

752 Tube Testing

Tubes were probably the most critical of all
components both from performance and manu-
facturing viewpoints. Proper performance of
the oscillator was based primarily on the char-
acteristics of the triode. The required response
of the amplifier was intimately related to prop-

sharp high-voltage pulses were observed on a
long-persistent screen oscilloscope. Improve-
ments in tubes and changes in operating point
of the oscillator apparently eliminated these
pulses and no further study was necessary. Ro-
tational frequency noise, that is, noise associ-
ated with the speed of the rotating system, was
caused by eccentric coupling shafts rotating in

Figure 15. Precessing ring vibrator.

the r-f field, gear trains which had binding
action at a certain spot, etc. The power supply
was an occasional strong source of rotational
noise, particularly when radio frequency was
present in the generator and gear train hous-
ing.

7 5 COMPONENT TESTING

Introduction

Proper performance of the major subassem-
blies of the fuzes depended on the careful selec-
tion of the various components. Close toler-
ances were set on the performance require-
ments of the subassemblies and these could only

COMPONENT TESTING

291

Table 1. Summary of diode tests.6* 35

Name of test

Limits and requirements (NDRC specifications,
Purpose of test Aug. 1, 1944)

Filament current*

Leakage test*

Self-noise test*

Rectified current
test*

Centrifugef

Operation!

To insure that filament current will
be within limits for satisfactory
operation.

To measure reverse d-c current
through leakage paths in the tube.

To measure a-c component of leak-
age current mentioned above.

To measure the diode efficiency, that
is, the ratio of the d-c voltage de-
veloped across a specified load to
the peak input voltage.

To determine ability to withstand
setback encountered in rocket fuze
application.

To insure adequate life expectancy.

Filament current shall be at least 60 ma d-c, and not
exceed 80 ma d-c, with 0.6 v d-c applied directly to
filament.

Leakage current shall not be greater than 3 na d-c
under operating conditions.

Maximum peak self-noise shall not exceed 0.1 mv.

Diode current shall be at least 30 n a d-c with 30 v
rms, 60 c, input.

To pass critical tests after acceleration of 2,500 g.

After operation for 15 min, the rectified current shall
not differ more than 10% from the value at
beginning of operation.

* These tests were given to all tubes (100% tests).

t These tests were given to representative samples from each lot of tubes (sampling tests).

Table 2. Summary of triode tests.6* 53

Name of test

Limits and requirements (Ordnance Dept.
Purpose of test specification, July 25, 1945)

Heater current*
(filament current)

Gas test*

Oscillation* (grid
bias test)

Oscillation frequency
testf

Self-noisef

Microphonics*

Operation testf

Centrifugef

To insure that filament current will
be within limits for satisfactory
operation.

To detect presence of gas which
creates fluctuations of plate cur-
rent and internal impedance.

To insure sufficient grid bias volt-
age can be developed to maintain
uninterrupted oscillation.

To insure that oscillator frequency
controlled by interelectrode tube
capacities will be held within
proper limits.

To measure spontaneous noise
within the tube.

To insure minimum noise voltage
when tube is subjected to vibra-
tion.

To insure satisfactory oscillation
performance after a period of
operation in excess of the ex-
pected time for combined testing
of the tube and the completed
fuze.

To determine ability to withstand
setback encountered in rocket and
mortar fuze application.

Filament current shall be between 0.230 amp d-c and
0.150 amp d-c with applied voltage of 1.20 v d-c.

Grid current must not exceed 2 na d-c between 1 and
6 sec after application of test voltages.

Grid bias of not less than 22 v d-c must be self-
developed with plate current of between 7 and 11
ma d-c for from 1 to 30 sec after application of
test voltages.

Frequency of each tube must not differ by more than
±5 me from standardized value.

Total instantaneous noise of tube at rest shall not
exceed 0.01 v.

The total instantaneous noise after amplification in
a shaped amplifier (gain approx 100) shall not
exceed 0.6 v when vibrated (0.012 in amplitude)
at a frequency in the pass band of the amplifier.

After 15-min operation under specified conditions,
oscillator grid bias must not differ by more than
20% from value noted prior to test.

Grid bias and plate current must not differ by more
than 20% from values before centrifuging, 12,000#.

* These tests were given to all tubes (100% tests).

t These tests were given to representative samples from each lot of tubes (sampling tests).

292

LABORATORY TESTING OF FUZES

important properties, particularly for the tri-
ode, was microphonic stability. This property
was examined on tubes before they were built
into fuzes. It was also examined, though in-
directly, in stability tests on completed fuzes
(see Section 7.4).

check a large percentage of the resistors and
capacitors to obtain some idea of the effect of
component variations upon fuze performance.
The production line, however, did not require
such information, but did find it necessary to
check certain critical resistors and condensers

Table 3. Summary of pentode tests.6- 53

Limits and requirements (Ordnance Dept.

Name of test

Purpose of test

specification, July 25, 1945)

Heater or filament
current*

Voltage amplifica-
tion*

To insure that filament current will
be within limits for satisfactory
operation.

To insure that voltage amplification
of pentodes is sufficient.

Filament current shall be between 0.072 amp d-c
maximum and 0.052 amp d-c at 0.6 v d-c.

Voltage amplification in a specified amplifier (no
feedback) shall not be less than 90 nor greater
than 120.

Noise*

(microphonics)

To insure that pentodes used in Total instantaneous noise, expressed as the maximum
fuzes are not excessively micro- peak variation in the plate voltage caused by any
phonic. single mechanical shock, shall not exceed 0.75 v

when the tube is subjected to the proper test
conditions.

Dynamic input im-
pedance!

Plate resistance!

To insure that input impedance is
sufficiently great.

To insure that plate resistance is
sufficiently great, since low plate
resistance affects the phase shift
of the feedback network.

The input impedance shall not be less than 10
megohms under operating conditions.

The plate resistance while the tube is operating in
the specified circuit shall be in the range 2.0 to
5.75 megohms.

Special (low) volt-
age amplification!

Operation test!

Mechanical stability
of elements!
(centrifuge test)

Surface electric
leakage!

To insure that the amplifier would
function properly with reduced
voltages.

To insure satisfactory voltage
amplification of the pentode after
a period of operation in excess of
the expected total time required
for testing the tube and com-
pleted fuze.

To determine ability to withstand
setback encountered in rocket and
mortar fuze application.

To insure that surface leakage of
the tube is not low enough to im-
pair its operation.

The voltage amplification when determined in the
specified manner shall not be less than 75.

After 15-min operation under specified conditions,
voltage amplification must not differ by more than
10% from value noted prior to test.

To withstand acceleration of 12,000# under certain
specified conditions, and 2,500# under other con-
ditions. The value of voltage amplification after
centrifuging shall not differ by more than 10 per
cent of the value prior to centrifuging, and the
noise after centrifuging shall not exceed 0.83 v
peak.

The minimum electric resistance between the plate
lead and all other leads shall be 25 megohms.

* These tests were given to all tubes ( 100% tests ) .

t These tests were given to representative samples from each lot of tubes (sampling tests).

A summary of the important tests on the
various tubes is presented in Table 1 for diodes,
Table 2 for triodes, Table 3 for pentodes, and
Table 4 for thyratrons.

Resistors and Capacitors
It was necessary for pilot line production to

to maintain a high level of fuze performance.
When such testing was required, ordinary com-
mercial-type limit bridges were used, although
several automatic sorting devices were pro-
posed and tried. Special surge testers to de-
termine the inductance of the cylindrically
wound detonator-firing capacitor were devel-
oped in order to design more efficient noninduc-

COMPONENT TESTING

293

Table 4. Summary of thyratron tests.6* 53

Name of test

Purpose of test

Limits and requirements (Ordnance Dept,
specification, July 25, 1945)

Heater or filament
current*

Critical grid voltage*

Grid circuit voltage
drop*

Minimum surge*

Constancy of critical
voltagef

Operation test part 1,
heater lifef

Part 2, repeated
surge!

Mechanical stability
of elements!
(centrifuge)

Internal electric
leakage

Surface electric
leakage!

To insure that filament current will
be within limits for satisfactory
operation.

To insure that critical grid bias is
within proper limits, since this
parameter is one which governs
overall sensitivity.

To insure that there is no excessive
leakage between the tube leads;
such leakage tends to make the
negative bias at the grid itself
lower than the applied C bias.

To insure that the minimum peak

• discharge current passed by the
thyratron will be sufficient to fire
the electric detonator.

To determine the effect of changes
in supply voltages on the critical
grid voltage.

To insure that the thyratron will
not undergo any deterioration in
operating characteristics after a
period in excess of the expected
time required for testing the tube
and completed fuze.

To insure that critical grid voltage
does not change appreciably after
the thyratron has been fired ten
successive times.

To determine ability to withstand
setback encountered in the rocket
and mortar fuze application.

To insure that electric leakage
within the thyratron is not ex-
cessive.

To insure that surface electric leak-
age about the thyratron is not
excessive.

Filament current shall be between 0.100 amp d-c and
0.080 amp d-c at 1.2 v d-c.

The critical grid voltage shall be within the range
— 1.7 to — 2.5 v d-c.

The grid circuit voltage drop shall not exceed 0.40
v d-c.

The peak plate current shall not be less than 5 amp.

Variation in the critical grid voltage with specified
changes in heater and plate voltages shall not
exceed 0.6 v d-c.

The tube shall be electrically stable as evidenced by
freedom from changes in critical grid voltage ex-
ceeding 0.40 v d-c after being operated for 15 min.

The tube shall be electrically stable, as evidenced by
repeated compliance with the minimum surge test,
freedom from changes in critical grid voltage ex-
ceeding 20% of initial value and from changes in
grid circuit voltage drop exceeding 0.20 v d-c, after
operation for 15 min.

To withstand acceleration of 12,000# under certain
specified conditions, and 2,500# under other condi-
tions; after centrifuging the critical grid voltage
shall not differ by more than 10% of pre-
centrifuging value, the value of grid circuit volt-
age drop shall not differ by more than 0.15 v d-c
from the precentrifuging value, and the minimum
surge current shall be greater than 4.5 amp.

The thyratron shall not pass more than 2.5 na d-c
when the control grid is tied to the plate and minus
135 v d-c applied.

The minimum electric resistance between the plate
lead and all other leads shall be 25 megohms.

* These tests were performed on all tubes (100% tests).

t These tests were performed on representative samples from each lot of tubes (sampling tests).

tive condensers. The reader is referred to refer-
ence 32 for the background of component tests
and specifications.

7,5,4 Coil Testing

In general, testing of oscillator coils con-
sisted of visual inspection to insure that ade-
quate cement had been applied and that the
proper number of turns had been wound. How-

ever, for transverse antenna fuzes (T-51) a
double coil was used. This double coil had to be
tested for high-voltage shorts or breakdowns.
Figure 17 shows a small test set which was used
for breakdown testing of coils and condensers.

5 Rectifier Assemblies

Since the rectifier buttons used in the fuzes

' SECRET

294

LABORATORY TESTING OF FUZES

were developed specifically for the fuzes, new
production test equipment was required to
make 100 per cent tests. This equipment was
designed to measure the forward voltage drop
and back current under specified conditions. In
addition, a probe was developed which would

Figure 17. High potential leakage tester.
Double-wound coil for T-51 fuze is shown being
tested for insulation between coils.

apply the proper voltages and compression or
contact force. Requirements for rectifier test-
ing are given in Section 3.4. 5.33’ 52

Chokes and Transformers

The small r-f filter choke was very delicate
by virtue of the very fine wire coil. This fine
wire would frequently be broken while bend-
ing the choke leads during assembly. A con-
tinuity test was, therefore, incorporated in the
audio prepot test position as an oscillator func-
tioning test. The audio chokes and transform-
ers used in certain fuzes (T-51 and T-82) were
frequently tested before assembly in mockup
amplifiers to determine their resonant fre-
quency. A mockup circuit was most satisfac-
tory, since it gave direct comparison data and
since direct inductance measurements of these
particular chokes were difficult.

7.S.7 Propeller and Turbine Assemblies

As mentioned previously, every effort to
eliminate vibration of the fuzes was made. One
of the largest sources of vibration energy came
from unbalanced rotating systems. Unbalanced
propellers and off-center coupling shafts pro-
duced considerable vibration in the large fuzes,
while dynamic unbalance of the turbine and
generator rotor shook the miniature fuzes.
Tests to determine unbalance were standard
production line procedure and the methods are
discussed in Sections 4.6 and 6.4.

•

Generators and Power Supply

In testing generators it was necessary to de-
termine that the bearings ran smoothly and
that the rotor would not break at high speeds.
It was also necessary, of course, to see that the
developed A and B voltages met specifica-
tions.7-34 Tests were made with the generator
working into a mockup of a typical rectifier-
filter section.

This mockup rectifier-filter section consisted
of four half-wave vacuum-tube rectifiers with
series resistors to match the forward resistance
of an average rectifier assembly and a parallel
resistor to match the average leakage resist-
ance (see Figure 32).

Tests on completed power supplies were
made while working into a typical load. Gen-
erators or power supplies not attached to fuzes
were driven by air turbines or high-speed
motors.

Adjustment of the A and B voltages devel-
oped by the generator was accomplished by
overmagnetizing the rotor and demagnetizing
to the proper amount while the generator was
under test. Overmagnetization followed by par-
tial demagnetization was found necessary to
obtain stable magnets.12

Several successful demagnetizing schemes
were used and one may be found in refer-
ence 42.

At this time it would be well to describe
briefly the types of meters used in measuring A
voltages. The waveform of the A voltage was not
quite sinusoidal because of transformer action

SPECIAL TESTS ON COMPLETED UNITS

295

between the A and B windings, which mixed a
square wave voltage, caused by the rectifier
load, and a sine wave voltage from the A wind-
ing which was under a resistive load. It was
therefore necessary to measure its rms value
as an indication of its effective, or heating
power. Ordinary rms meters which used either
a thermocouple, dynamometer movement, or
moving vane were unsatisfactory for routine
tests. The thermocouple meters were sensitive
to ambient temperature variations, would re-
spond to stray audio and r-f voltages, were
very sensitive to overload, and in general were
of such low impedance that they applied an
additional indeterminate load on the power
supply. The dynamometers had very low im-
pedance, requiring power equal to or greater

voltage source whose waveform matched that
of a loaded generator. Calibrated thermocouple
voltmeters were used as standards.

Meters whose resistances were 5,000 ohms
per volt or greater were required to measure
B and C voltages in order not to load the power
supply.

For measuring the speed of rotation of gen-
erators, tachometers were designed that were
actuated by the frequency of the A voltage. In
general, the tachometers consisted of an ampli-
fier, wave clipper, and vacuum-tube voltmeter
which read the average voltage across an RC
net work. A number of circuits were developed
for this purpose8-22 but the most satisfactory
one, as well as the one used most extensively,
is shown in Figure 18.

Figure 18. Circuit diagram of tachometer used for measurement of rotational speeds.

than the total generator output. The moving
vane meters were extremely frequency de-
pendent and vefy insensitive at frequencies
greater than 500 c. A vacuum-tube voltmeter
was developed for laboratory use but was un-
satisfactory for routine testing. It was demon-
strated that the waveform between generators
of a given type were sufficiently uniform to
allow the use of rectifier-type meters, provided
these meters were frequently calibrated with a

7 6 SPECIAL TESTS ON COMPLETED UNITS

76,1 Introduction

Although the observations in the final test
position (referred to in Section 7.4 and de-
scribed in detail in Section 7.8) provided the
most important data on completed fuzes, other
special or supplementary tests were required.
The nature of these tests varied with the type

SECRET

296

LABORATORY TESTING OF FUZES

of fuze. Some were 100 per cent or production
tests, that is, they were performed on all fuzes
of a particular type. Others were sampling
tests ; that is, only representative samples from
production lots were tested.

The most important of the special tests on
completed fuzes are described in this section.

762 Pulse Test

The pulse test was given each fuze after final
assembly in its encasing can to insure that it
was operating electrically after all test leads
had been disconnected and the encasing can
staked. The pulse test occurred after the “final”
test, since the latter was made with a number
of special test leads soldered to the fuze. An
r-f pulse of appreciable magnitude was im-
pressed on the fuze by grounding a metal plate
near the nose, or grounding the antenna itself.
If all electric circuits were continuous and
functioning, the thyratron would fire a simple
neon firing indicator connected to the detonator
contacts in place of a detonator. The pulse test
thus provided a simple overall check of the
assembled unit in its encasing can.

7 6 3 Tests of Self-Destruction Circuits

For those devices with self-destructive [SD]
circuits (T-5), tests were performed to check
the length of time required for this circuit to
function. Both mechanical and electric SD de-
vices were used. The SD time could be roughly
measured with a stop watch. However, circuits
were devised which recorded the time on an
electric clock.

Arming Pulse

An arming pulse test was performed on T-5
fuzes to insure that transient pulses due to the
application of voltage to thyratron plate and/or
disturbances of the r-f field present at the arm-
ing switch did not fire the thyratron.5 If the
thyratron did not fire when armed, the fuze
was satisfactory in this respect. The test was

not considered necessary on later models (T-50,
T-51, etc.), since extensive laboratory tests of
these designs demonstrated the absence of arm-
ing pulses.

765 Warmup

A warmup test was performed as a sam-
pling test on T-5 units to insure that initial cir-
cuit transients were below firing magnitude
when arming occurred. The transients were
primarily due to filament warmup and con-
denser charging.5

766 Apex Firing

An apex firing test was under development
for mortar fuzes, but such a test was never
actually performed in pilot production testing.
The necessary investigations were made, how-
ever, for the test which was intended to insure
that the thyratron would not fire at the apex
of the trajectory where the supply voltages
developed by the generator dropped to very low
values. A number of factors occur at the apex
of the trajectory which tend to fire the thyra-
tron either as the generator slows down or
speeds up.2G

76 7 RC Arming Delay

The RC arming delay was measured on fuzes
which used this type of arming as an additional
safety feature. The test insured a certain mini-
mum capacitance for the detonator-firing ca-
pacitor and an RC product within specified
limits. Details of these requirements as they
pertain to RC arming are given in Section 3.3.6.

768 Shelf Life

Shelf or storage tests were performed at
periodic intervals on a group of T-5 fuzes by
subjecting them to the usual performance
tests.31 The most serious effect of long storage
was detuning (cf. Section 3.1).

SECRET

SERVICE TESTS

297

" 7 SERVICE TESTS

7,7,1 Introduction

The service use of fuzes in wartime involved
conditions of transportation, handling, storage,
and installation which imposed severe require-
ments on design and construction. Ruggedness
and resistance to extremes of atmospheric con-
ditions were essential properties of the fuzes if
they were to withstand rigors of wartime use.
Some of the extreme conditions involved in
transportation and storage were mollified by
the container or package in which the fuze was
shipped, but the fuze was still subject to con-
siderable rough handling and atmospheric ex-
tremes after unpackaging. The bomb fuzes, for
example, were required to be carried in bomb
bays for extended periods at very low tempera-
tures and then to operate properly when re-
leased.

The so-called service tests described in this
section were designed to test the ability of fuzes
to perform properly under operational condi-
tions.

7,7 2 Jolt Test

Fuzes were subjected to a jolt test (on a
sampling basis) to test mechanical strength
and ruggedness of construction. The test was
performed according to Ordnance Department
specifications. The jolt machine used for the
test consisted of a series of arms, operated by
rotating cams, which held the fuzes (Figure 37,
Chapter 4) . As the cams rotated, the arms were
raised in turn and allowed to drop on a wooden
block. It was required that the fuzes pass criti-
cal operating tests after jolting.

Vibration and Packaging Tests

A vibration test was performed on fuzes to
simulate conditions resulting from vibration
of an airplane in flight. The test was adapted
from the Navy Department Bureau of Ships
specification for type testing of airborne elec-
tronic equipment. The test consisted of vibrat-

ing fuzes at frequencies of from 10 to 55 c at an
0.06-in. amplitude for 30 min. It was required
that the fuzes pass critical operating tests after
vibration.

Standard ordnance packaging tests were per-
formed on fuzes in their container. The tests
included shock, jumbling, and exposure to
atmospheric extremes. The fuzes were tested
for overall performance before and after the
packaging test, and if the fuzes performed
properly on retest, the packaging was consid-
ered satisfactory. The packaging tests, except
for the electric measurements, were made at
Picatinny Arsenal.

7,7,4 Temperature Tests

Two types of temperature tests were made
on fuzes, one on the completed fuze and the
other on the head and the power supply. The
first, a temperature cycling test, was made on
the completed fuze and simulated the alternate
extremes of high and low temperature en-
countered in transportation and storage. The
fuzes were subjected to a number of cycles of
high and low temperatures, after which they
were tested for mechanical and electric per-
formance. After the test, it was required that
a fuze meet certain electric and mechanical re-
quirements which allowed limited changes from
prior performance. The second test, which was
performed on the fuze head and the power
supply separately, was an operational test per-
formed while the head, or power supply, was
actually operating under a condition of extreme
temperature. The temperatures used were —40
and +60 C. It was required that certain param-
eters of the head, or power supply, should not
differ by more than certain small percentages
from their values when measured at a tempera-
ture between 20 and 30 C. (See specifications
listed in the bibliography of Chapter 5.)

7 7 5 Humidity Tests

Humidity tests were made in order to dupli-
cate conditions to which fuzes would be sub-
jected in tropical climates where during the day

SECRET \

298

LABORATORY TESTING OF FUZES

there was high humidity and temperature, with
lower temperatures and high humidity at night.
Tests were made in a controlled humidity cham-
ber where the temperature could be cycled,
duplicating the breathing as occurs in service
use. As with the temperature cycling tests,
performance data after humidity cycling had
to be within certain specified limits of the prior
values. (See specifications listed in the bibli-
ography of Chapter 5.)

7'7'6 Salt Spray Tests

Salt spray tests were made to determine the
effect of the corrosive action of sea water and
spray. Tests were made according to Army and
Navy aeronautical specification AN-QQ-S-91,
with the requirement that the fuzes operate
satisfactorily both electrically and mechanically
after treatment.

7’7’7 Centrifuge or Accelerating Tests

Centrifuging tests were performed to deter-
mine the effects of appreciable acceleration on
the fuzes. Only those fuzes which would experi-
ence accelerations in service were so tested
(rockets and mortars). The chief defects
caused by centrifuging were failure of mechan-
ical parts and displacement of circuit elements
which created changes in electric performance.
Both commercial and specially built centrifuges
were used. The smaller fuzes could be accommo-
dated in commercial centrifuges, but a special
double-beam type of centrifuge was built for
the larger fuzes4 (see Figures 32 and 33, Chap-
ter 4) .

7/7 8 Field Test Set IE-28

A field test equipment known as the IE-28
test set was developed for field testing major
subassemblies of T-5 fuzes. It is shown in Fig-
ure 19. It was designed primarily to test bat-
teries before final assembly in the field but was
also arranged to provide tests for the arming
switch and the electronic assembly (MC-382).

The test of the latter was a pulse test similar
in purpose to the one described in Section 7.6.2.
Tests on the switch checked the safety (i.e.,
unarmed condition) and continuity of the elec-
tric detonator.

Simple laboratory tests of this sort, made in
the field just prior to use, were considered de-
sirable for T-5 fuzes primarily because of the
newness of the fuze as an ordnance item. No
similar tests were considered necessary or de-
sirable for generator-powered fuzes.

The IE-28 test set was made to test either T-5,
T-6, or T-4 (photoelectric) fuzes. The fuze head
(MC-380) shown in Figure 19 is part of the T-4
fuze.b

Figure 19. Field test set IE-28 with T-4 fuze
in position for testing.

7 8 MECHANICAL TESTS AND GAUGING

781 Introduction

Mechanical tests were required to insure
maximum safety of the fuzes and proper opera-
tion of the mechanical parts, particularly the
high- and low-speed rotary systems. Gauging
operations were performed on dimensions

b This fuze is described in Division 4, Volume 3,
Summary Technical Report.

SECRET

MECHANICAL TESTS AND GAUGING

299

which were critical in determining interchange-
ability of parts or in determining fit or clear-
ances in Service use.

7 8 2 Static Torque Tests

A static torque test was made on the wind-
mill to determine the torque required to turn
it from a stationary position. The gauge used
incorporated a spring device which either indi-
cated the torque directly on a scale or caused a
light to glow if the measured torque was
greater or lower than specified limits. The pur-
pose of the lower torque limit was to insure the
existence of sufficient magnetic lock (see Sec-
tion 3.4.5). Windmills of fuzes meeting this
requirement would not turn below a certain
minimum air velocity (about 150 fps). The
purpose of the upper torque limit was to insure
that the rotary system was free to turn.

The static torque test was repeated with the
fuze under compression28 and at various tem-
peratures between —40 and +60 C. This test
was made to insure free turning of the rotary
system when subjected to the force produced
by tightening the booster cup and when oper-
ated under extreme conditions of temperature.
The compression applied during the torque test
was sufficient to give an indication of the com-
pression strength of the fuze. The force was
applied (for ring-type fuzes) between the in-
terrupter plate and the antenna ring.

A torque test was performed on the detona-
tor rotor to insure that the force required to
move this rotor into its final position would not
be too great on account of possible stiffness in
the detonator contact springs. The test was per-
formed with a torque gauge similar to that used
for the static torque test for the windmills. It
was found that adherence to the limits of this
test was an important factor in preventing
duds.

78 3 Binding and Dynamic Torque Tests

A mechanical binding test was performed
on the completed fuze to insure that no tight
spots existed in the rotating system due to
tight or defective parts. If the speed of the fuze
was within certain limits when driven by a low

and constant pressure airstream, it was con-
sidered satisfactory. This same test was per-
formed under temperatures ranging from —40
to +60 C to insure that mechanical binding
would not occur at extreme operating tempera-
tures.

A dynamic torque test was made on the
larger type fuzes, i.e., bomb fuzes, where the
propellers could be driven by a motor drive.29
The purpose of the test was to insure that the
dynamic torque required to drive the rotating
system would be within limits which would not
give undue variations in arming times. Too
little or too great dynamic torque would cause
higher or lower propeller speeds respectively,
with corresponding variations in arming times.
The torque was measured by means of a tor-
sion wire, the torque reaction being measured
by the amount of twist of the wire (Figure 20) .

Figure 20. Torsion wire dynamometer used to
measure dynamic torque.

The torques measured were of the order of 1.3
in.-oz at 8,000 rpm.

7 8 4 Other Mechanical Tests

Dipole strength tests were made on bar-type

SECRET

300

LABORATORY TESTING OF FUZES

fuzes by applying a force at a point % in. from
the outer end of the dipole and perpendicular to
both the axis of the dipole and the axis of the
fuze. The original test called for an 80-lb
applied force, while 150-lb force was later
specified for models which used stronger
plastic materials in the nose. The requirement
for the test was that the dipole should suc-
cessfully withstand several applications of the
force.

785 Gauging

Gauging was performed on dimensions which
were critical in determining interchangeability
of parts. Thread gaugings were probably the
most important operations. In the complete
fuze assembly three sets of threads were in-
volved, namely, threads on the casting contain-
ing the electronic assembly which mated with
threads in the encasing can (potato masher) ;
outside threads on the encasing can for screw-
ing the fuze assembly into the missile; and
threads on the tetryl cup which mated with
threads in the encasing can. These threads were
gauged with appropriate thread gauges. Other
types of dimensions gauged included diameters
and depths of holes and overall lengths of parts
and threads. Ordinary commercial snap, plug,
sight, and concentricity gauges were used, as
well as many special gauges developed par-
ticularly for the jobs at hand.

In addition to gauging the critical dimen-
sions mentioned above, measurements were
made of electric and mechanical arming angles
and the height of the detonator contact springs.
These three items were critical because im-
proper adjustment of any one or all of them
could cause improper operation of the arming
system. Electric arming angles were measured
with an automatic turns counter. Mechanical
arming angles and contact spring height were
measured with suitable gauges.

Vane blade angles of metal windmills were
measured with a Bausch and Lomb comparator.
These measurements were important in keep-
ing the effective pitch of the windmill constant.
Variation in pitch would, of course, cause vari-
ations in time to arming (see Section 9.2.2).

Bakelite windmills were not subject to such
variation since they were molded.

The spring tension of the transfer pin con-
tained in the detonator rotor was measured to
insure its proper function in springing out to
release the rotor from the slow-speed shaft and
then hold it in the armed position. Insufficient
spring tension might permit the detonator
rotor to ride beyond the armed position and
cause a dud, while too much tension would drag
the shaft with a possible failure of the gear
train, again causing a dud or excessive drag
on the generator. It was also necessary to check
the alignment of the transfer pin with respect
to the keyway of the slow-speed shaft and the
arming hole wires; incorrect alignment might
produce binding of the rotating system.

Mechanical life tests were not usually run
except during experimental or pilot production.
The procedure used in such cases was to sub-
ject fuzes to mechanical operation for a given
length of time and then test them electrically
to determine any changes from previous per-
formance. This process was then continued in-
definitely until mechanical or electric break-
down occurred or until it was apparent that the
fuzes under test had more than satisfactory
mechanical life.

7 9 PILOT PRODUCTION TEST LINE

791 Introduction

During the process of unit assembly it was
desirable to make certain routine checks to in-
sure the completion of high-quality fuzes and
a low percentage of rejects. The proper testing
of units during assembly prevented systematic
difficulties (lots of poor components or im-
proper assembly operation) and permitted re-
pairs while components were accessible.

Test positions which combined the essential
laboratory tests and techniques outlined above
and yet would not appreciably slow down the
assembly process were, therefore, designed to
check assemblies at various times during pro-
duction.

When designing test equipment for the model
shops and pilot production lines the most diffi-

SECRET

PILOT PRODUCTION TEST LINE

301

cult problem was incorporating the special fea-
tures required by the various types of units so
that new equipment would not be necessary for
every fuze model. This requirement led to the
development of universal test equipment. That
is to say, equipment was developed in which
rewiring of the test panels was not required
every time a new fuze type was to be tested.
The panel contained switches or plug-in assem-
blies which could easily be altered as the test
specifications required.

In addition, when possible, the equipment
was “unitized,” that is, made of component
assemblies which could be used interchangeably
at different places and could be easily replaced
in case of failures. For example, the universal

numerable test jigs were designed and used,
since the various manufacturers had prefer-
ences in techniques. Some stressed simplicity,
some ease of operation, some accuracy, etc., but
essentially they held the item to be tested and
were similar to those illustrated below.

Figure 21 shows a typical test line, which,
it will be noted, follows the assembly procedure
outline in Figure 1 of Chapter 6.

Subsequent Figures 22 to 39, inclusive, show
pictorial diagrams and schematic circuits illus-
trating the test positions of Figure 21. The
illustrations are for one type of fuze ; however,
modification of the fixtures and voltage dividers
was all that was required for the other fuze
types, except for OD and POD fuzes. For the

Figure 21. Pilot shop production test line. Drawing references are listed in the Bibliography.

test panel and cathode follower were used in
five of the ten positions, and the tachometer in
three positions.

Since this equipment was to be used continu-
ously and had to yield data of considerable
accuracy, prime design factors were simplicity,
ease of operation, ruggedness, and precision.
Simplicity and ease of operation could not be
overstressed, because complexity confused op-
erators (who were relatively unskilled) and
made maintenance and calibration difficult. In-

latter the circuit was arranged to measure plate
current instead of grid voltage. For the OD
fuzes an extra switch position was required for
reading diode voltage. (These models were not
in production at the close of World War II;
therefore, the remainder of this section will
deal with the RGD type fuze. References 2, 5,
40, and 51 contain details for testing specific
fuze types.)

The presentation in the remainder of this
Section follows the block diagram of Figure 21.

SECRET

302

LABORATORY TESTING OF FUZES

7 9 2 Oscillator Pretest Position0

In the oscillator pretest position, the r-f
assembly was complete except for the antenna
ring, cap, or dipoles, since the antenna was not
added until just prior to the head test position.
The subassembly was mounted on jigs consist-
ing of a shield can which contained a properly
adjusted resistance and reactance load to com-
pensate for the absence of the antenna. An elec-
tronic power supply replaced the generator for
the operating voltages. It might be pointed out
that any voltage equivalent to 1.4 volts rms was
satisfactory for the filaments, so that for sim-
plicity a filament transformer in the power line
was used. A milliammeter in series with the B
voltage measured the plate current, while a
high-resistance voltmeter read the grid voltage
developed by the oscillator. Any of the vari-
ous types of wave meters (or receivers) could
be used to measure the frequency, provided it
was not too tightly coupled to the oscillator. In
general, there was sufficient radiation from the
test leads to operate the wave meter. Otherwise
a probe was inserted into the shield can.

For pilot and model shops it was desirable to
use directly calibrated meters and indicators,
since the data acquired were used for correla-
tion purposes. However, production line equip-
ment was frequently marked so that the
indicators showed only the tolerance limits.

Audio Pretest Position*1

In the amplifier pretest position, the sub-
assembly included the amplifier and thyratron
circuits. Here measurements were made of the
millivolts to fire, and when necessary, the fre-
quency shaping and the normal critical voltage
of the thyratron. The gain-adjusting gimmick
was set at this test position (in designs which
incorporated this feature). The millivolts to
fire were measured at either the amplifier peak
frequency or at fixed frequencies to determine
the characteristics of the pass band.

Supply voltages were obtained from regu-
lated electronic supplies which were set to de-

c See Figures 22 and 23 and drawing reference 1.

d See Figures 24 and 25 and drawing reference 2.

liver prescribed voltages (1.4 volts at 1,000 c,
140 volts direct current, 7.5 volts, and 2 volts
direct current) to the amplifier under test. This

Figure 22. Oscillator pretest position.

position contained the necessary thyratron fir-
ing indicator, cathode* follower, and output in-
dicator, audio voltage source, and voltage

Figure 23. Schematic of oscillator pretest posi-
tion.

divider. The voltage divider was in the form of
a plug-in assembly which could easily be
changed when testing other types of amplifiers.

PILOT PRODUCTION TEST LINE

303

This position consisted of the following unit-
ized panels: (1) regulated power supply, (2)
universal test, (3) audio oscillator, and (4)
the necessary jigs.

For those fuzes which had the rectifier and
filter circuits housed in the audio compartment

Figure 24. Audio pretest position.

(T-82, T-132, T-171, T-172), power supplies
were provided to furnish a-c voltage to the rec-
tifier-filter network. In the latter cases, it was
possible to make a spot check of millivolts to
fire with these components in operation.

794 Audio Prepot and Postpot Test
Positions6

After the oscillator and amplifier were con-
nected together and assembled into the chassis
(casting), quick checks of oscillator grid volt-
age, plate current, millivolts to fire, and normal
critical voltage were made to insure no errors
had been made in assembly. Then, after the
assemblies were potted, they were again
checked to eliminate those units whose charac-
teristics had changed abnormally during the
potting process. Certain small systematic
changes were expected because of the change
in stray capacitance caused by the added dielec-
tric material.

These test positions were similar to the audio
e See Figures 26 and 27 and drawing reference 3.

pretest except for the addition of a grid voltage
meter and a jig which contained an r-f load
similar to that used in the oscillator pretest
position.

7 9 5 Head Test Position£

When a fuze assembly reached the head test
position, it was completely assembled except
for the power supply. The antenna cap, ring, or
dipole (depending on the type of fuze) had
been sealed in place. Here any final adjustments
were made,g such as final setting of the gain
control gimmick and determining the value of
the padding resistor which normalized the B
load of the headed unit. Normalizing of the B
load was essential to keep the C bias within
specified limits (see Section 3.4.5).

The test panel was identical to the audio post-
pot panel except for the jig.

The r-f jig consisted of a 2-ft shield box con-
taining the necessary mount and lead connec-
tions. An r-f load (as described in Section
7.2.4) matching the free-space load was con-
sidered part of the r-f jig. Sensitivity and
stability tests were made at this position as
required.

7 9 6 Generator Test Position11

The generator was checked not only for the
A and B voltages developed across specified
loads, but also for alignment of rotor and pole
pieces and the ability of the rotor to withstand
high speeds. With the newer rotors, the latter
becomes less important. Heavy mechanical
shielding was necessary, however, to confine
rotor fragments if one should fracture.

The A voltage was measured with conven-
tional a-c voltmeters when the proper resistive
loads were applied across the windings (see
Section 7.5.8). The B voltage was rectified and
filtered with a mockup vacuum-tube rectifier
which was adjusted to duplicate characteristics
of an average rectifier-filter section. In general,

f See Figure 28 and drawing reference 4.

s It should be mentioned that at this test position, the
diode resonant circuit of OD units was tuned and locked.

h See Figures 29 and 30 and drawing reference 5.

SECRET

304

LABORATORY TESTING OF FUZES

Figure 26. Audio prepot and postpot test posi-
tion.

Figure 27. Schematic of audio prepot and post-
pot test position.

SECRET

PILOT PRODUCTION TEST LINE

305

the voltages were measured at specified speeds,
and occasionally regulation data was obtained
at this test position. Regulation and B/A ratio
tests were not performed as routine tests and
usually required precision circuits (see Section
3.4.5).

Figure 28. Head test position.

While at this position, the rotor was de-
magnetized until the proper voltages were
obtained or until a slightly higher voltage was
obtained. The latter procedure permitted

Figure 29. Generator test position (Bowen).

further demagnetization at a later test posi-
tion to compensate for differences in loads and
rectifier-filter performance.

Rectifier-Filter Test Position1

The rectifier-filter section consisted of a
* See Figures 31 and 32 and drawing reference 6.

resistor-condenser network which rectified and
filtered the a-c B voltage. Taps were included
for the required C-bias voltage or voltages. In
addition, this assembly usually contained the
contacts for the detonator rotor.

An a-c voltage was supplied to the rectifier-

Figure 30. Schematic of generator test posi-
tion. Rla and Rlb — load resistors adjusted as
specified. X and Y adjusted to match ideal
rectifier assembly.

filter section from a generator whose output
impedance was similar to that of the fuze
generator. The a-c voltage required to produce
a given d-c voltage across the output and the
no-load source voltage were measured. These

Figure 31. Rectifier filter test position (Bowen).

two values give an indication of the efficiency
of the rectifier subassembly.

Metering circuits for measuring the C bias,

SECRET

306

LABORATORY TESTING OF FUZES

the a-c ripple across the B load, and the con- air turbine driver, contact prods, and the A
tinuity of the detonator leads were included. and B loads.

Power Supply Test Position3

Final Test Position11

After assembly of the generator and rectifier-
filter, the completed power supply was tested

Figure 32. Schematic of rectifier filter test posi-
tion. Mi — 0-400 volts alternating current, M2
— 0.200 volt direct current and 0-10 volt alter-
nating current. Filter section to see a total load
(Rl) of 8,000 ohms ± 2 per cent including meter
resistance. Effective impedance of source looking
back from points X-X shall be 7,700 ± 1 per
cent including R and meter resistance. Reactive
component shall be less than 1,000 ohms.

and the voltages adjusted for the proper fuze
load. This final adjustment was accomplished
either by demagnetizing the rotor or loading
the B filter and adjusting the C-bias network.

Figure 33. Power supply test position.

The test panel contained the necessary A, B,
and C voltage meters, a tachometer, and de-
magnetizing equipment. The jig contained an
j See Figures 33 and 34 and drawing reference 7.

The performance of a fuze at final test was
established as the basis for product acceptance.

tion.

For this reason, this test position was most
elaborate. Here the critical electric voltages
were measured (A, B, and C, oscillator grid
bias, millivolts to fire, and effective critical
voltage) under conditions which simulated as
near as possible free-flight conditions. The free-

Figure 35. Final test position for ring-type
fuzes.

space r-f load was duplicated in a shielded box
or chamber, as described in detail in Section
7.2.4. The fuze was mounted on a mechanical
k See Figures 35, 36, 37, and 38 and drawing reference

SECRET

PILOT PRODUCTION TEST LINE

307

system which permitted vibration, while the
circuits were operated by power supplied by
the fuze generator. The generator was driven
with a stream of high-velocity air directed
through suitable jets at the vanes of the wind-
mill.

Figure 36. Final test position for bar-type
fuzes (Zenith).

The most difficult problem connected with the
final test position was the securing of adequate
contact to the test points and obtaining the
proper vibration as discussed in Section 7.4.1.
With the standard size fuzes, test leads were
soldered to the connecting lugs on the amplifier
base plate, as no other system was devised
which permitted good contact under the neces-
sary conditions of vibration. (See fuze on table
in Figure 6.) In designing the miniature fuzes,
special terminal boards (side view, Figure 14B)
were incorporated which made possible the use
of small, quick-acting clamps on which were
mounted all the necessary test prods.

The test panel consisted of a tachometer and
universal panel with the voltage meters, audio
input and output circuits, and effective critical
voltage measuring equipment.

The fixture was a shielded box or metal
chamber containing the r-f load, air jet, and a
resonant vibration mount similar to those de-
scribed in Section 7.4.2. and Figures 2 and 5.

High-velocity airstreams were the most prac-
tical method of driving the windmills or tur-
bines. The air was obtained from a line

(pressure at 80 to 100 psi) and directed at the
vanes through jets made from dielectric ma-
terials. The first jets were made of glass (see
Figure 39). Plastic jets soon replaced these
since the mortality of glass jets was very high.
In places where r-f loading was not important,
metal jets were used (see Figure 51, Chapter
4). The jet assembly consisted of an air
reservoir from which vents issued. These vents
were directed almost normal to the vane sur-
face with a slight incline toward the leading
edge to direct the airflow beyond the propeller.
Typical plastic jets are shown in Figures 5,
14B, and 15. Jets of this type consumed approx-
imately 20 cu ft per min.

Figure 37. Inside view of final test position for
bar- type fuzes (Zenith).

7 9 10 Pulse Test1

The pulse test was the last test made on a
fuze before packaging. The fuze was at this
point complete except for explosive elements.

1 See Figure 39 and drawing reference 9.

SECRET

308

LABORATORY TESTING OF FUZES

All the test leads had been removed, and the
encasing can staked in place. The purpose of
the test was simply to insure that during these
final operations no connections were broken or
no leads short-circuited which would prevent
the fuze from operating. Removing the test

Figure 38. Schematic of final test position.

leads and staking the cans presented oppor-
tunity for accidents.

The complete fuze was mounted in a shielded
box to eliminate extraneous noises and driven
with an air blast. A neon lamp in series with a
protective resistor and by-passed by a con-
denser and resistor were mounted on a deto-
nator rotor in place of the detonator to serve
as a firing indicator.

An r-f disturbance was produced to fire the
thyratron by grounding an r-f pickup plate or
by grounding the fuze antenna. The fuze was
considered satisfactory if the neon fired only
when the r-f field was disturbed.

7 10 QUALITY CONTROL TESTING

710 1 Object of Quality Control

Quality control laboratories were set up to
check and control54 the quality of all types of
fuzes which reached the production stage. The
accuracy of testing equipment at these labora-

tories was necessarily high. In addition, they
served as a calibration center for checking
equipment at the manufacturers. It was the
express function of the quality control labora-
tory to indicate the trends of test data, note
defects of manufacture and failure to meet
specifications, and immediately to report the
results to the manufacturer.

Quality control testing was under control of
the Services; Division 4’s participation in it
was primarily of an advisory nature. Develop-
ment of the test equipment used for quality
control was, however, done largely in Division
4’s central laboratories at the National Bureau
of Standards. The intimate relationship be-
tween design of the equipment and the operat-

Figure 39. Pulse test position.

ing properties of the fuzes made development
of test equipment an integral part of the fuze
development program.

Test equipment used in the quality control
laboratory was the same as used on similar
tests on the production line. The production
line, however, included tests on some parts and
subassemblies which were not duplicated at
quality control.

As mentioned in the introduction to this
chapter, the sequence of operations at quality
control was, in general, the reverse of those
used on the production line. Quality control

SECRET

QUALITY CONTROL TESTING

309

began with a completed fuze and broke it down ;
the production line started with parts and sub-
assemblies which were assembled into a com-
pleted fuze. Another difference was in the
nature of the data taken. Actual values were
recorded at quality control, whereas limit
meters were usually used in production.

Eighteen typical fuzes from each lot of 1,000
were submitted to the quality control labora-
tory. Later, as quality improved, smaller
numbers were submitted. These samples were
selected at random from regular production by
a representative of the contracting officer sta-
tioned at the factories. Tubes submitted to
quality control were selected in a similar
manner.

The routing of fuzes in the laboratory is
shown in the following diagram (Figure 40)
which is discussed briefly. The discussion will

mentally or on fuze models which had not
reached the production stage.

Quality control tests were of three general
classes: (1) production, (2) sampling, and (3)
type. The first were performed on all fuzes in
the sample (the same tests were also presum-
ably made previously by the manufacturer).
Sampling tests were made on only two or three
fuzes from each sample lot. These tests were
more difficult to perform and impaired the
quality of the fuze so that it was no longer
suitable for field use. Type tests were usually
of the same general nature as sampling tests,
but were made only at irregular intervals. Par-
ticular occasions for making type tests were

(1) on the qualification lot of a new design,

(2) when a change in production procedure
had been made which might conceivably change
the properties of the fuzes, and (3) when

Figure 40. Flow diagram for quality control.

show only the position which each test occupied
in the quality control program, since a detailed
discussion of each has been included earlier in
this chapter. Tests mentioned previously, but
not included here, were used only experi-

production tests had indicated an undesirable
trend in the product.

In the following discussion the letters P, S,
and T are used to indicate whether the tests
were production, sampling, or type. ST indi-

310

LABORATORY TESTING OF FUZES

cates type tests which were often run as
sampling tests.

7.10.2 Mechanical Tests and Gauging

Upon being received and checked in, a lot of
fuzes was routed to a laboratory where mechan-
ical and gauging tests were performed. These
tests included

1. Gauging of mechanical arming angle (P).

2. Measurement of propeller turns to electric
arming (T).

3. Gauging height of detonator contact
springs (S).

4. Gauging of detonator rotor housing (P).

5. Gauging of detonator rotor and transfer
pin (P).

6. Measurement of tension of spring in
transfer pin; alignment of transfer pin (P).

7. Gauging of tetryl cup and tetryl plate (P) .

8. Static torque test (P).

9. Detonator rotor torque test (S).

Tests on gear trains were run on a sampling
basis on the product used by the fuze manu-
facturer. Gear trains from finished fuzes were
not tested separately.

Following mechanical tests and gauging, a
visual inspection (P) was made for the purpose
of noting possible mechanical defects, poor
workmanship, or extraneous materials.

710 3 Overall Electric Tests

While the fuze was in the mechanical labora-
tory, in its encasing cans, it was given the
pulse test (P). (See Section 7.6.2.) The encas-
ing cans were then removed from all fuzes in
the lot and the lot sent to the electric laboratory
where the test leads were soldered in place.
There the fuze was given the following electric
tests and the mechanical binding test (final
test position, Section 7.9.9).

1. A, B, and C voltage (P).

2. Diode voltage for OD fuzes (P).

3. Oscillator grid bias voltage (P).

4. Oscillator plate current for POD fuzes (P).

5. Carrier frequency (P).

6. Peak audio frequency (P).

7. Millivolts to fire at one or more frequencies (P).

8. Effective critical voltage and/or noise margin test

(P).

9. Normal critical voltage (P).

10. Thyratron grid circuit voltage drop (P).

11. Mechanical binding test (P).

In connection with certain of the final per-
formance tests, it should be pointed out that
data for curves of millivolts to fire versus audio
frequency, curves of effective critical voltage
and C volts versus generator speed, and curves
of A and B voltage versus generator speed may
be made at the final test position. The milli-
volts-to-fire curves indicate the frequency
characteristic of the amplifier, the intersection
of the effective critical and C voltage curves
show the lowest speed at which the generator
can operate without premature firing of the
thyratron, and the A and B voltage curves in-
dicate the regulation characteristic of the gen-
erator. Such data were usually taken on a
sampling or type basis.

7 104 Head Tests

Following the final performance test, part of
the lot was given the head test which consisted
of the following.

1. Tuning for OD fuzes (S).

2. Oscillator stability test (ST) .

3. R-f sensitivity test (ST).

4. Oscillator plate current test (ST).

All head tests on a group of units were usually
performed by one operator, though not neces-
sarily at one test position. An external voltage
supply was used to furnish power (usually
alternating current for the filaments and direct
current for the plate supply).

710 5 Special Tests

After the head test, various fuzes of the lot
were routed to the remaining tests included on
the diagram, namely,

1. Destructive mechanical tests.

a. Dynamic balancing tests (ST).

b. Strength tests (S).

c. Vane pitch measurements (S).

2. Jolt test (ST) .

3. Vibration test (ST).

SECRET

QUALITY CONTROL TESTING

311

4. Humidity cycling test (ST) .

5. Temperature cycling test (ST).

6. Extreme temperature performance tests (ST).

7. Salt spray tests (ST).

The strength tests included dipole strength
tests (T), for bar-type fuzes, and compression
tests (ST) on the complete fuze assembly.

7 10 6 Tube Tests

Tubes were supplied to the fuze manufac-
turers by the Army. In order to maintain the
quality of the tubes as furnished, the Army
also required that quality control tests be run

on tube production. The tests performed were
essentially those outlined in Section 7.5.2.

7.10.7 Limits and Tolerances

The limits and tolerances of performance re-
quired in quality control were somewhat arbi-
trary. In general, a compromise was made be-
tween ideal performance and allowances neces-
sary to maintain production. Figures 8, 9, 18,
and 19 in Chapter 6 indicate for some param-
eters the degree of uniformity that was ob-
tainable. The limits imposed for the various
fuzes are given in the specifications listed in
the bibliography of Chapter 5.

SECRET

Chapter 8

FIELD TESTING OF PROXIMITY FUZES’

81 GENERAL INTRODUCTION

Throughout the development of radio
proximity fuzes for nonrotating projectiles,
field tests were relied upon to provide informa-
tion under conditions which, because of igno-
rance of what such conditions were or because
of the difficulties involved, could not be dupli-
cated in the laboratory. It was demonstrated
repeatedly that there was no laboratory sub-
stitute for field testing. From the first field tests
in the early part of 1941, in which the charac-
teristics of the reflected signal were investi-
gated, to the service acceptance tests, field tests
gave vital data on conditions under which the
fuzes must operate, the effects of electrical and
mechanical changes in design, and the quality
of the final service fuzes.

Experience gained in the field testing of prox-
imity fuzes also provided the basis for the nec-
essary special instructions for handling the
fuzes in service use. Although the special pre-
cautions for handling the fuzes (due primarily
to the fact that missiles became radio antennas)
were simple and easy to perform, they were
essential for the best performance of the fuzes.

For the performance of these tests, proper
proving ground facilities for releasing bombs
from planes and for firing rockets and mortar
shells had to be provided, and technical meth-
ods, mainly electronic, photographic, and pyro-
technic, had to be devised or adapted to provide
the desired information. At first, when the vol-
ume of field tests was small, the facilities of
established proving grounds were used, but
later, as the volume increased, most of the test-
ing of proximity fuzes was done in new proving
grounds or ranges set up or reserved for the
sole purpose of testing proximity fuzes. How-
ever, special tests continued to be performed in

n This chapter was prepared by Theodore B. Godfrey
of the Ordnance Development Division of the National
Bureau of Standards. David C. Friedman of the same
organization assisted in the preparation of parts of
Section 8.2. Section 8.4, including photographs and
diagrams, was taken almost in its entirety from Chap-
ter 7 of the Final Report of the University of Iowa.42

other locations, particularly tests on effect fields
and demonstrations performed at the request
of, or in cooperation with, particular branches
of the Armed Forces.

The first field tests in connection with the
development of bomb fuzes were performed in
the early part of 1941 at Camp Springs, Mary-
land, the Naval Air Station, Lakehurst, New7
Jersey and the Naval Proving Ground, Dahl-
gren, Virginia. The greatest part of subsequent
bomb fuze testing was performed at the Aber-
deen Proving Ground, although important tests
were also performed at Dahlgren, Virginia,
Eglin Field, Florida, and Edgewood Arsenal,
Maryland. At Aberdeen alone some 14,000
bombs were dropped in the field testing of bomb
fuzes.

In February of 1942, rocket fuze tests were
started at the Aberdeen Proving Ground. The
need for additional facilities was soon evident,
and a new proving ground was established at
Fort Fisher, North Carolina, where test firing
started in May 1942. Special tests, particularly
those in which high-explosive [HE] loaded
rockets were used, continued to be made at the
Aberdeen Proving Ground. In the early part
of 1943, a second proving ground, devoted to
tests of proximity fuzes alone, was established
at Blossom Point on the Potomac River, about
40 miles south of Washington, D. C. At first,
because of the presence of a commercial air lane
over this proving ground, firing was restricted
to low elevations, but later in the year, after
the air lane had been shifted to the west, per-
mission to fire at any elevation was obtained.
In December 1943, the proving ground at Fort
Fisher, where approximately 11,000 rounds had
been fired, was abandoned. At the Blossom
Point Proving Ground nearly 14,000 rocket and
mortar rounds were fired up to September 1,
1945.

The firing of mortar shells in connection with
the development of the mortar shell proximity
fuzes was started at Blossom Point in April
1944 and at the Clinton Field Station near
Clinton, Iowa, in May 1945. Up until Septem-

312

BOMB TESTS

313

ber 1, 1945, approximately 1,950 mortar rounds
were fired at Blossom Point and approximately
1,700 rounds at Clinton.

Since the methods used in the developmental
field testing of bomb, rocket, and mortar shell
fuzes were often the same or very similar, some
repetition has resulted in order to make Sec-
tions 8.2, 8.3, and 8.4 each reasonably complete.
When more details than are given in one par-
ticular section on some particular method are
desired, they will often be found in one of the
other sections. Details of proving ground tech-
nique are treated most fully in Section 8.4 on
mortars; the described methods of testing are
more numerous in Section 8.3 on rockets but
there is less detail.

The test procedures described are, in gen-
eral, those employed in tests performed under
direct supervision of Division 4 (at Blossom
Point, Aberdeen, and Clinton, particularly).
Safety precautions, except as they were peculiar
to proximity fuzes, are not discussed. The
Safety and Security Division of the Ordnance
Department, through periodic inspections, made
recommendations which were very helpful in
reducing the hazards connected with field test
operations. Precautions for proper installation
of the fuzes to secure best performance are
discussed. These precautions were incorporated
in Service manuals issued by the Ordnance De-
partment for use with the fuzes.

8 2 BOMB TESTS

8,2 1 Introduction

Field tests of bomb proximity fuzes gen-
erally involved mounting the fuzes on standard
bombs and dropping the fuzed bombs from
standard bombers over the area in which ob-
servations were made. Both inert and HE-
loaded bombs were used.

The great majority of bomb fuze tests under
the direct supervision of Division 4 were per-
formed at the Aberdeen Proving Ground. Fig-
ure 1 as a graphical representation of the accu-
mulative total number of bombs dropped at
Aberdeen.

Experiments with inert bombs were made

to obtain information on the following points.

1. Fuze reliability.

2. Height of function.

3. Fuze generator characteristics (speed,
speed regulation, bearing performance).

4. Arming distance.

Experiments with HE-loaded bombs were
made to obtain information on the following

Figure 1. Accumulative number of bombs
dropped at Aberdeen.

points in addition to those listed for inert
bombs.

1. Minimum separation in train required to
eliminate mutual interference.

2. Fragmentation pattern.

3. Optimum height of burst.

4. Effectiveness of air burst in comparison
with ground burst.

822 Bombs Used

Standard bombs were generally employed in
field tests and included those listed in Table 1.
The weights given are nominal.

314

FIELD TESTING OF PROXIMITY FUZES

In addition to these standard bombs, other
vehicles used have included the T-15 and T-16
fragmentation bombs, various fighter fuel tanks

Table 1. Bomb types.

Type

Designation

Weight

(lb)

General purpose

M-30

100

M-57

250

M-64

500

M-65

1,000

M-66

2,000

Semi-armor-piercing

M-58

500

Light case

M-56

4,000

Fragmentation

M-88

220

M-81

260

Incendiary

M-76

500

Chemical

M-47

100

M-70

115

M-78

500

M-79

1,000

modified to carry napalm, and the British
4,000-lb blast bomb.

Preparation of Bombs

The location of the bomb at the time of func-
tioning of the fuze was obtained by photo-
graphic methods. Therefore, the preparation of
HE-loaded bombs for tests of proximity fuzes
presented no special problems, since the high
explosive provides its own photographic flash.
Other than the observance of standard pre-
cautions in the handling of HE-loaded bombs,
care had only to be taken to insure that the
fuze and tail were tightly mounted, since ex-
cessive mechanical vibration or intermittent
electric contacts might cause random function
of the fuze. Secure mounting of the fuze and
fin to the bomb was recommended as standard
procedure in the Service use of the fuzes.

The use of inert bombs, however, is to be pre-
ferred in developmental testing. The handling
of bombs and the taking of data can be done
much more conveniently without the hindrance
of precautions which must be observed when
high explosives are used. Therefore, a spotting
charge arrangement to indicate the location of
fuze functioning was devised for use with inert
(mainly sand-loaded) bombs.

Spotting Charge. For the composition of the
spotting charge itself, a mixture of 76 per cent

200-mesh potassium permanganate and 24 per
cent 200-mesh magnesium (proportions by
weight) was found to be very satisfactory. This
mixture was loaded into cardboard cartridges,
5 in. long and 1 in. in diameter. (The prepara-
tion of such spotting charges is extremely
hazardous and should be performed only by
professional fabricators of pyrotechnical de-
vices.) The spotting charge was taped to the
rear end of the fuze, one end of the cartridge
being in contact with the tetryl-filled booster
cup.

It was found that the explosion of the booster
and spotting charge was not capable of ejecting
the fuze from the nose of a sand-filled bomb.
Therefore, a steel pipe, 2 in. in inside diam-
eter, was used to conduct the flame from the
spotting charge to the tail of the bomb. An
empty bomb case was used; the bottom of the
fuze seat liner was knocked out and the tail
plate removed. A piece of steel tubing, cut to
the proper length, was inserted into the bomb
case from the tail end and its forward end
slipped into place over the fuze seat liner. The
tail end of the tubing was plugged and centered
by means of a special funnel through which the
bomb was filled with sand. The funnel and tube
plug were removed and the tail plate installed.
The tail end of the tube was then centered and
held in place by a tail fuze adapter from which
the bottom had been cut out or by a special
plastic plug.

It was found that fragmentation bombs,
which have a low HE to total weight ratio,
could be used satisfactorily for proximity fuze
tests without sand loading and, therefore, with-
out the installation of the steel tubing. The
only preparation required consisted of knock-
ing out the bottom of the fuze seat liner and
removing the tail plug.

With this arrangement the interval of time
between explosion of the detonator and the ap-
pearance of flash at the tail of the bomb is of
the order of 4 msec.12’ 15 After the appearance of
the flame a dense green smoke, easily visible to
the eye, is formed.

8 21 Assembly of Fuze Components

Fuzes were received for field test minus all

BOMB TESTS

315

explosive components. The detonator rotors
were loaded at the National Bureau of Stand-
ards with the use of M-36 detonators (tempo-
rary developmental designations, T-3-E1,
BS-5) . The rotors were checked on a special test
box for detonator continuity and transfer pin
operation ; only a small fraction of the minimum
firing current was used in the detonator test.
The rotors were inspected for projecting det-
onator contacts and any other roughness which
might hinder operation.

The final assembly was done at Aberdeen
Proving Ground. First, the rotor was inserted,
the relative positions of the keyways in the
shaft and rotor housing furnishing a rough
check on the setting for minimum safe air
travel. Next, the tetryl-filled plate was dropped
in and engaged on the projection which held it
in proper alignment. Then the tetryl booster
cup was screwed in hand-tight. If the fuze were
to be used on an inert bomb, the spotting charge
in a cylindrical cardboard tube (1 x 5 in.)
was secured to the cup with a short piece of
2-in. Scotch Tape.

8 2 5 Assembly of Fuze to Bomb

The fuze and puff combination was assem-
bled to the bomb, either in the bomb bay or
before loading of the bomb into the bay. The
fuze was screwed in hand-tight when the
spring-type washer was used and hand-tight
plus J/2 turn with a wrench when the lock
washer was used.

The arming arrangement was examined to
make sure that the arming wires would pull
out properly, that there were no kinks where
the wires might break under the load, and
that the arming pins would function properly
(cf. Figure 20, Chapter 4).

A wrench was used to tighten the fin lock-
ing nut, special care being exercised to elimi-
nate rattle and possible consequent early func-
tioning of the fuze.

Delayed arming devices, when used, were
checked for release and for free spin before
attaching, and for engagement and setting
after the bomb and fuze had been mounted in
the bomb bay.

826 Range Layout

Although some testing was done over
ground with cloth targets for aiming points,
and some over water with rafts as aiming
points, so many disadvantages became ap-
parent that a permanent water-target range
was laid out at Aberdeen. The diagram used
in locating impact points on this range is
shown in Figure 2. The land-water boundary,
not shown, lay close to the observers’ stations.

The two permanent targets were built on
piles. The horizontal surfaces (painted white)
were planked with 2xl0’s on 20-in. centers
and were about 10 ft above the water. Pile
clusters were driven about 15 ft from each
corner to protect the targets from drifting
ice. A 25-ft pole was erected on each target to
aid in estimating function heights.

The inner target, 1,980 ft from shore, was
30 ft square and was used for tests with inert
bombs. The outer target, 2,215 ft beyond the
other, was 20 ft square and was used for tests
with HE-loaded bombs. No fragments from
proper functions were ever observed to fall
closer than the inner target. Fragments from
random functions, however, fell 3,000 or 4,000
ft horizontally from the point of burst. There-
fore, the bombing course was laid over a
cleared area.

82< Communications

Plane-to-ground communication was accom-
plished with conventional Signal Corps re-
ceiving and transmitting equipment. The
bombardier gave a 5-sec warning of “ready”
and “fire” at moment of release. This was
audible at each station through a wired inter-
communication system. In addition, each sta-
tion was provided with a portable radio re-
ceiver, which could be used to hear signals
directly from the bombing plane.

8 2 8 Testing Conditions

Routine tests of fuze quality required the
determination of function score, function

f SECRET \

316

FIELD TESTING OF PROXIMITY FUZES

BOMB TESTS

317

height, and fuze operation in flight. Factors
influencing function score and function height
include fuze sensitivity, vertical component of
bomb velocity, angle of approach to target,
and reflection coefficient of the target. To re-
duce the number of parameters, bombing was
usually done at a standard speed (200 mph)
and from a standard altitude (10,000 ft) over
a large body of water with an essentially con-
stant reflection coefficient. From laboratory
tests and results of field tests under these con-
ditions, the performance under other condi-
tions could be computed.

Some tests were carried out under other
conditions of plane speed and altitude to
simulate, as far as fuze operation was con-
cerned, the conditions of dive bombing. It
proved very difficult to reproduce release con-
ditions in dive bombing. Accordingly a system
of release conditions was worked out for
horizontal bombing in which the striking
angles and approach velocities of the bomb
were the same as for various conditions of
dive bombing.27 Still other tests were made
from higher speeds and at higher altitudes to
test the units under more severe conditions,
or to test attachments for the fuzes.

829 Determination of Function Heights;

Visual Methods

Function heights were estimated visually
by a trained observer to make possible early
discussion of test results and to supplement
the photographic data, if incomplete. The ob-
server was aided by the 25-ft pole mounted on
the target and checked his estimates regularly
with the photographic heights. The differences
between visual estimates and photographic
heights, obtained by trained observers, were
remarkably small.

In addition, a camera obscura was often
used as a visual means of obtaining function
heights. Battery commander telescopes at the
north and south stations and triangulation
were used to obtain the range, so that the
apparent height could be converted to actual
height.

For engineering purposes, however, data ob-

SEC

tained by photographic methods were usually
required because of their greater accuracy.

8.2.10 Determination of Function Heights;

Photographic Methods

Data for photographic determination of
function heights were obtained by means of
16-mm motion-picture cameras, placed at the
north and south stations (see Figure 2). Koda-
chrome film was exposed at the nominal rate of
64 frames per sec and was slightly overexposed
in order to give the correct density for use on
Recordak Viewers. Usually, a 2-in. focal
length lens was used, but 15-mm, 1-in., and
3%-in. lenses were also available. The 1-in.
lens was used for photographing trains of
bombs or in other tests where a larger field of
view was required.

Normally, the camera was aimed at the
target, and pictures of function, target and
impact were obtained without moving the
camera. However, on some drops it was neces-
sary to swing the camera to obtain the picture
of the impact. In such cases the photographer
would note the azimuth of the splash with re-
spect to the target. This was obtained from a
scale mounted on the tripod head of the
camera support. Often functions of wild drops
could be photographed because the noise of
the bomb could be heard, or the bomb seen,
soon enough to swing the camera toward the
point of function.

Each photographer was provided with a
stopwatch, which he started when the bomb
release signal was heard over the intercom-
munication system. Knowing the time of flight,
he could start the camera a few seconds before
impact.

In order to identify each round the photog-
rapher took pictures of a data board, on which
were marked the station, the date, the photog-
rapher’s name, the lens used, and the round
number.

In order to provide an easy means of setting
the scale on the Recordak Viewers, the photog-
rapher would take a picture at the beginning
or end of the day’s work of a scale pole from a
distance of 500 ft, using the same lens as that

:et

318

FIELD TESTING OF PROXIMITY FUZES

used for the day’s program. The scale pole was
painted alternately black and white in foot-
wide strips. In addition, the first and last feet
were marked with horizontal boards, which
projected out from the pole. These bars showed
up, even when the lighting or exposure was so
poor that the black and white strips did not
show with sufficient clarity.

When the film was received in the computing
room, it was enlarged to a convenient size on a
Model 10 Recordak Viewer. The picture of the
target pole was used to set the scale, which
usually was 1 ft on the pole equal to 2 mm on
the screen.

The target was usually near the center of
the film and was considered to be exactly there
for purposes of computation. The error involved
by the assumption was usually small. The
horizontal distance from function to target and
the vertical distance from function to splash
were measured. If flash and splash were not
in the same frame, the water-land boundary or
the horizon were used as reference points. If
the camera had been swung, the irregularities
in the skyline were used as reference points in
going from frame to frame, or the cameraman’s
notes were used to find the azimuth of the
function.

Work was then transferred to a plotting
board (Figure 2), consisting of a scale drawing
of the range (1 mm equal to 10 ft) with scales
running through the target perpendicular to
the line between the camera and the target
(one scale for each camera position). Addi-
tional millimeter scales were pivoted at the
camera station points for locating the function.
The horizontal distance measured on the
Recordak, corrected for change in scale, was
then transferred to the scale on the board, and
the pivoted scale was passed through the pro-
jection on the horizontal plane of the apparent
point of function. The same would be done for
another camera station. Where the two scales
crossed would be the projection of the func-
tion on the horizontal plane. The distances
from the two cameras would then be read off
directly from the pivoted scales. The measured
vertical distances to the point of function on
the film were then converted to the true vertical
distances at the proving ground. This gave two

values for the function height. In order for the
determination to be acceptable, agreement be-
tween these two figures within 5 ft was re-
quired. The average was reported as the func-
tion height.

If the camera had been swung so that taking
the projected distance would lead to too great
an error, or when this distance could not be
obtained because of blurring, the pivoted scale
was passed through the azimuth obtained, as
described in previous paragraphs.

It is estimated that the position of the flash
may be determined by the method just de-
scribed within 5 ft of its true position. How-
ever, the position of the flash of the spotting
charge may not give the true height of function.
A later section discusses possible error due to
time lag of the spotting charge (see Section
8.2.13).

When very accurate function heights were
required, a three-dimensional analysis of the
film data was made.9 Two projectors were set
up to represent the cameras used in taking the
film. The lenses were in the same ratio as those
used to take the pictures. The range was laid
out to scale, and a screen with the target loca-
tion marked on it was set up at the scale dis-
tance to the target. The two projectors were
then started and the pictures run, until the
first sign of the spotting charge appeared. The
film was then stopped and the cameras pivoted
about an axis through the optical center of the
lens until the images of the target coincided
with the target location on the screen. The
screen was then moved forward or backward,
until the images of the spotting charge burst
coincided. This located the burst in space. The
apparent height of the burst was then measured
and the scale of the setup used to convert this to
actual function height. The accuracy of this
method depends upon the scale used, the exact-
ness of the match in lens ratio, and the ability
of the operator to superpose the images. This
method may be used to locate in space any por-
tion of the trajectory made visible by smoke
tracer or other means.

8211 Determination of Function Time

Function times were determined in several

BOMB TESTS

319

ways, depending upon the accuracy required.
The usual means was visual determination by
observers with stopwatches. The release signals
used also depended upon the accuracy desired.
The least accurate was a voice signal given by
the bombardier over the plane-to-ground radio.
The bombardier gave a 5-sec warning, then the
release signal, when he dropped the bombs. This
signal could be heard at the various stations
by means of the intercommunication system.
There were two types of automatic signals: a
photoflash bulb on the plane fired when the
bomb release was operated ; and/or a squegging
oscillator either started or stopped at release
which could be heard via radio and the inter-
communication system on the ground.

Function time was also determined from re-
cordings on film or phonographic disks, of the
output of a radio receiver tuned to receive the
r-f carrier of the transmitter in the fuze. The
starting signal, except when the squegging
oscillator was used, was a standard 1,000-c
note started as a warning by the chief observer
and stopped by him when he saw the photoflash.
When the bombs were dropped in train, this
observer also recorded the times of random
functions by momentary pulses of the 1,000-c
note at each observed function. The fuze carrier
was picked up by the receiver and its modula-
tion detected, passed through a limiter28 and
recorded. The end of the modulation, except in
rare cases, was an indication of the function or
of the impact. By matching the 1,000-c note to
a standard in the computing room, it was pos-
sible to run the record at the same speed as it
was in the field and to obtain several deter-
minations by stopwatch of the function time.
By running the record at half or one-third
speed, the error in time could be decreased.
When film records were used, the film was
driven by a synchronous motor, and a neon light
recorded a time scale along the edge of the
record of carrier modulation. When the line
frequency was known, the time from the end
of the 1,000-c note to the end of the carrier
could be determined. When the line frequency
varied too much, a 50-c oscillator31 was used to
operate the neon light.

Flight time was occasionally obtained by
photographing a clock and the function on the

same picture, using a high-speed camera. The
clock was started at the release signal.

In most of these methods an error arises be-
cause of the use of one observer’s reaction time
to start but not to end the recorded interval. It
has been estimated that the average error in
observing function times was about 0.2 sec.

8 212 Observations of Fuze Carrier
Characteristics

Investigation of fuze operation in flight was
accomplished by picking up the fuze carrier and
recording the modulation on film or phonograph
disks, or both. Depending upon the frequency
of the unit, a Hallicrafter S-27 or S-37, equipped
with a long wire antenna, was used to receive
the carrier. The audio output was passed
through a limiter stage28 to a Presto K-8 phono-
graphic recorder or to a Dumont 3-in. oscillo-
graph with a 16-mm camera attachment29 or
both. Speakers were also connected to the re-
ceiver output, one for the operator and one for
the chief observer. The starting signal, which
also provided the time scale signal, consisted of
a 1,000-c note, which was started as a warning
signal and stopped as a release signal by the
chief observer. Round identification and perti-
nent remarks were placed on the record by
means of a microphone connected directly to
the recorder amplifier. The radio operator kept
a list of film run numbers versus round numbers
for identification of film traces.

The strength of the fuze carrier, which gives
an idea of the overall performance of oscillator
and power supply, was read from an S meter
on the receiver.

The modulation may be divided into three
types: noise, microphonics, and ripples asso-
ciated with generator operation. The presence
of noise usually indicated some mechanical
trouble, such as binding or chattering gear
trains or bearings, grinding off of parts of gear
teeth, loose assembly of parts of the fuze or of
the bomb, or bits of metal in the generator.
Some of the troubles could be identified by com-
parison with records made in the laboratory.

Microphonics were caused by vibration of the
elements of the oscillator tube and indicated

SECRET

320

FIELD TESTING OF PROXIMITY FUZES

need for better mounting of the tube or better
support of the tube elements.

Generator ripples gave data on the perform-
ance of the vane or turbogenerator system. A
sudden change or an irregular variation in
ripple frequency usually indicated bearing
trouble. Some types of harmonic content indi-
cated rectifier failure.

Considerable reliable data on generator
speeds could be obtained from records of modu-
lation of the fuze carrier. However, when ac-
curate information on the speed characteristics
of new driving systems was desired, it was
customary to build special units each consisting
only of an oscillator, generator, and driving
system all in the regular fuze housing. In these
units the plate supply of the oscillator was ob-
tained directly from the generator, thus provid-
ing strong modulation. Such units were called
radio reporters.1 The radio reporters gave
superior data on generator speed because there
was less confusion as to the frequency of modu-
lation. In fuzes, modulation of the carrier by
the generator occurs due to filament modulation,
at generator frequency, and due to plate produc-
tion, at twice generator frequency.

Film records were first used to obtain gen-
erator speed data, but their use was later con-
fined to those cases in which greater accuracy
was necessary.

The film was placed on a Recordak Viewer to
enlarge the trace to a size convenient for count-
ing. Since a synchronous motor was used to
drive the camera, the distance between sprocket
holes could be used as a time scale. With these
cameras, when the line frequency was 60 c, the
time scale was %0 sec per space between two
adjacent sprocket holes. The line frequency dur-
ing the run could be found by counting the
number of cycles of the 1,000-c starting signal
per frame space at various points of the trace.
The average count was taken to be the value to
be used for the run. On well-regulated power
lines this scale was very accurate and con-
venient to use. When the line frequency was not
constant enough, a 50-c tuning fork oscillator
was used to light a neon bulb, which left a
series of dots along one side of the film. This
could be used as the time scale.

The generator speed at a given time was ob-

tained by counting the modulation frequency
over a short time interval, centered about the
time in question. This frequency could then be
converted to generator revolutions per minute
as follows :

If the modulation was caused by filament
ripple, as was usually the case, speed was deter-
mined by the formula

c X 60
6 t X n*

where s = generator speeds in revolutions per
minute,

c = number of cycles counted in time
interval,

t = time interval in seconds, and
n = number of pairs of poles in gen-
erator.

If the modulation frequency were due to
plate ripple from a rectified power supply, a
factor of 2 would appear in the denominator,
since full-wave rectification doubles the gen-
erator ripple frequency.

Because of harmonic content in the modula-
tion, it was often difficult to know whether the
generator voltage frequency was being counted
or a multiple or submultiple of that frequency.
Therefore, for accurate work on film, reporter
units, in which the modulation was deliberately
enhanced, were used.

Because film work was slow and hard on the
eyes, phonographic methods of determining
generator speed were developed. It was also
found that the phonographic method would
yield data in the case of units which had con-
siderable noise and microphonic modulation,
which would ordinarily make film work im-
possible.

Two methods were generally used. In the
first, a good record was essential, as any har-
monic content tended to cause a frequency lower
than the true one to be recorded. The record
was played back, and the output of the player
was led to a General Radio frequency meter,
whose output was led through a General Radio
d-c amplifier to a recording milliammeter. The
trace was a direct frequency versus time curve.
Corrections were necessary to make up the
time differences caused by a difference in line
frequency between the place where the record

SECRET

BOMB TESTS

321

was made and where it was played. The 1,000-c
starting signal was used to calibrate the mil-
liammeter scale.

The second and most used method of deter-
mining generator speed13 depended upon the
ability to match a note on a record with that
from an oscillator, either visually or aurally.
Because the usual record of carrier modula-
tion had a fairly rapid change in frequency
(since the bomb and hence the generator were
accelerating) , accurate frequency determina-
tion could not be made from the original
record. A re-recording was made, therefore,
through an electronic switch,30 which allowed
the output of the recorder amplifier to reach
the cutting head only at definite intervals,
usually about 0.8 sec, and only for a very short
time, usually about 0.2 sec. This recording
sounded like a record of a series of distinct
monotones, gradually changing in frequency.
If the frequencies were very high, the original
record would be played at half speed, while the
re-recording was made. The new record was
then played on a turntable, whose speed could
be adjusted and played at full or half speed. The
1,000-c note was used for setting this speed,
and the matching note was obtained from a
standard oscillator in the radio building of the
National Bureau of Standards.

The output from the record was led to a
loudspeaker and one pair of plates of an
oscillator (see Figure 3). The output of an
audio oscillator (Hewlett-Packard 200B) was
led to a single crystal earphone and to the other

input switch panel output

Figure 3. Block diagram of apparatus for de-
termining generator frequencies.

pair of oscilloscope plates. The record was then
allowed to play until a note was reached of
which the frequency was to be determined. Here
a stop kept the tone arm from traveling farther,

and, because of the spacing of the notes, one
note only would be played over and over again,
as the arm hit the stop, jumped back a groove,
and came to the stop again. The frequency of
this note was matched by varying the fre-
quency of the audio oscillator, until the two
notes sounded the same and no beat notes could
be heard. A Lissajous figure on an oscilloscope
was used to improve the accuracy of match. An
exact match would be indicated by a stationary
figure, but since the frequency of the note on
the record was changing, even though it
sounded like a monotone, the attempt was
made not to get a motionless figure, but one
which moved only slightly.

When the frequency had been determined, the
record was played through from the beginning.
The time from the starting signal to the match-
ing note was obtained by a stopwatch, several
determinations being made. Since the fre-
quency could be changed into generator speed
very easily in a manner similar to that ex-
plained already, the use of this procedure led
to the determination of points on a generator
speed versus time curve. By the use of bomb
velocity versus time curves, these data could
be translated into bomb velocity versus gen-
erator-frequency data, and vane slip factor data
could be obtained. This method was rapid and
sufficiently accurate for engineering purposes.
In the case of noisy or microphonic records, the
ear could discriminate between the desired and
the extraneous frequencies where an instru-
ment could not.

While the ear could not determine whether
the frequency was the direct generator voltage
frequency or some multiple or submultiple, the
general range of speeds could be obtained
from film or phonographic records of reporter
units on which the frequencies could not be
mistaken.

8,2,13 Special Tests

Arming Tests

Because safety and arming data are of great
importance to the users of variable-time [VT]
fuzes, there had to be devised tests from which
it would be possible to obtain data on the arm-

SECRET

322

FIELD TESTING OF PROXIMITY FUZES

in g time and arming distance. Since the bal-
listics of the bombs were known, the problem
reduced to that of finding accurately the time
to arm of many units with various arming set-
tings on various bombs.

Two ways of indicating arming were used.
In the first, the unit was modified to func-
tion on arming, so that ground observation of
arming time would be sufficient. The ways of
obtaining function time, mentioned previously,
then applied to these special tests.

The second way consisted of using units
which would function normally but with modi-
fications, such that the carrier modulation would
be changed in some way at completion of me-
chanical arming. Reporter units could also be
arranged in this way. A reporter unit, in which
the transfer pin did not spring out of the slow-
speed shaft and in which the plate circuit was
shorted when the rotor contacts were in the
armed position, would indicate arming by an
interruption of the carrier, which lasted until
the rotor contacts had turned out from under
the stationary ones. If the filter condenser were
not grounded until arming, an otherwise normal
unit would indicate mechanical arming by a
sudden cut off in modulation, followed by modu-
lation at one-half the former frequency. With
a short RC delay incorporated in the arming
system, this shock would not trigger the fuze,
which could then ride through to function on
the target.

The use of T-2 arming delay units made this
problem more complex, since safety considera-
tions prohibited the use of VT fuzes set to func-
tion immediately after the operation of a de-
layed arming device whose arming character-
istics were unknown. Such devices prevented
the unit from emitting a carrier, until after
the arming device had dropped off and the
warmup period was over.

Since the fuze had to arm normally, even
though shortly after the release of the T-2
device, it was possible from ballistic data, the
generator speed versus time curve, and the arm-
ing setting of the fuze to determine the time at
which the delayed arming device dropped off.
The fuze was usually set to function on arm-
ing.

A special unit could have been used for very

accurate determination of release of the de-
layed arming device, if more accuracy had been
necessary. Such a unit would consist of a
dummy fuze with an oscillator built into it, so
arranged that the oscillator would be cut off at
the release of the delayed arming device. A
laboratory test would provide the field crew
with the carrier frequency of the oscillator, so
that it could be tuned in immediately, allowing
the delayed arming device to be set for short
aiming periods. The determination of arming
time would then be the same as that of function
time, previously described.

Train Tests

In order to test the mutual effects of VT-
fuzed bombs, several tests were made, in which
bombs were dropped in trains with various in-
tervalometer settings. Some of these were made
with HE-loaded bombs, others with inert-loaded
bombs. In most of these tests, one or more of
the fuzes in each train was set to function on
arming and after the other fuzes had armed.
In this manner it was made certain that there
would be at least one early function, and the
stability of the other fuzes in the train could
be observed.

Data on the time of early functions were ob-
tained in several ways. The chief observer, who
handled the release switch, would put short
peeps of the 1,000-c note on the record of car-
rier modulation by momentarily pressing the
switch at each function. He also used several
stopwatches to obtain the times of such func-
tions. A camera was also mounted in the plane,
and the frames from the time of the first func-
tion to those of the other functions were
counted. The relative function times could be
obtained from this information as the rate at
which the film was exposed was known.

Such film would also show the relative posi-
tion in the horizontal plane of any functions
occurring close together. A follow-down camera
would show these from another angle.

Proper functions over the water were some-
times hard to locate, because the smoke or flash
from one would hide another, and, in the case
of HE-loaded bombs, the curtain of spray
thrown would obscure the impact positions. In
such cases it was often necessary to rely on

SECRET

BOMB TESTS

323

visual estimates of the range of function heights
for a closely spaced train.

Dive Tests

In order to obtain data on the performance
of VT fuzes in dive bombing, some tests were
made in which bombs were released at various
speeds and at various angles of dive. A camera,

Figure 4. Accumulative number of rockets fired
from stationary launchers.

in which a focal plane shutter was operated by
hand while a rotating shutter was continuously
driven by a motor, was used to photograph the
path of the plane. The variations in plane speed,
dive angle, and point of release led to the
abandonment of this method in favor of simu-
lated dive tests. In these tests a pilot flew a level
course at such an altitude and speed that the
approach of the bomb to the target was the
same as it would have been if the bomb had
been released under certain dive bombing con-
ditions. The impact angle of the bomb was ob-
tained by the three-dimensional analysis method
already mentioned. The use of a high-speed
camera, combined with the others, made pos-
sible the determination of terminal velocities.27

Determination of Time Lags

One of the questions which arose during the
testing program was whether the spotting
charge appeared at the same place as where
the fuze function occurred. Static tests were
made in which a photoflash bulb of known time

Figure 5. Accumulative number of rockets
fired from airplanes.

characteristics was fired at the same time the
detonator was fired. The explosion of the spot-
ting charge was photographed by a high-speed
camera, and the time between firing and the
appearance of the flash was obtained.12 It was
found that the time taken for the explosion to
travel down the tube from the fuze to the tail
of the bomb was of the same order of magni-
tude, 5 msec, as the time for the bomb to travel
its own length at the usual release conditions
and function heights, so that the flash would be
approximately where the fuze was at function.
It is not known how the speed of puff travel is
affected by the bomb velocity. It is also not
known whether the HE burst would occur
exactly where the spotting charge burst does.

secret

324

FIELD TESTING OF PROXIMITY FUZES

Figure 6. Range for high-angle rocket firing
at Fort Fisher.

The position of the spotting charge function is
close enough to the expected position of an HE
burst to be used for all but a few limited appli-
cations of the fuze.

8 3 THE FIELD TESTING OF ROCKET FUZES

Introduction

Although radio proximity fuzes for rockets
were developed largely for air-to-air or air-to-
ground use, most of the experimental data re-
quired during development could be, and were,
obtained by firing fuzed rockets from launchers
located on the ground or on ground-supported
towers. As required, these data were supple-
mented by data obtained from tests, some of
which were quite extensive and which were per-
formed at Aberdeen, Dahlgren, Eglin Field,
and Inyokern, in which rockets were fired from
airplanes. In both types of firing, a large pro-

Figure 8. View of east and west ranges,
Blossom Point.

portion of the rockets were inert except for a
spotting charge to indicate functioning of the
fuze.

Chronologically, but with considerable over-
lapping, the largest volume of firing of rockets
under the direct supervision of Division 4,

Figure 9. Smaller concrete magazines at
Blossom Point.

NDRC, was done first at the Aberdeen Proving
Ground, then at Fort Fisher, North Carolina,
and finally at the Blossom Point Proving
Ground, which was located on the Maryland
shore of the Potomac River about 40 miles
south of Washington. In Figure 4 accumulative

Figure 7. View of Blossom Point, looking west
from the firing tower.

Figure 10. Large concrete magazine at Blossom
Point.

THE FIELD TESTING OF ROCKET FUZES

325

curves of rocket rounds fired from stationary
launchers at the various proving grounds are
shown. Similar curves for rounds fired from
airplanes, not including those fired at Eglin
Field and Inyokern in tests conducted by the
Armed Services, are given in Figure 5.

ward the west and northwest from the top of
the firing tower at Blossom Point. Figure 7
shows the main group of buildings and Figure 8
shows the west range and part of the east
range. A tow target is suspended between the
poles of the west range. The poles on the right

Figure 11. Rocket ranges at Blossom Point.

Figure 6 is a photograph of the range used
for high-angle firing of rockets at Fort Fisher.
The main laboratory building appears at the
bottom of the picture and the balloon hangar
at the top. As indicated by dust clouds, a rocket
had just been fired from a mobile launcher
when this photograph was taken. The beach
which appears in the upper left-hand corner is
on the Atlantic Ocean.

Figures 7 and 8 are photographs taken to-

are on the east range, which was used for the
test firing of photoelectric fuzes. The three con-
crete magazines at Blossom Point are shown in
Figures 9 and 10.

A map of the Blossom Point range is shown
in Figure 11. The directions of fire often used
are indicated, together with the locations of
navigation lights which were often used as
reference points in night firing.

Details of instruments used in visual and

SECRET

326

FIELD TESTING OF PROXIMITY FUZES

Figure 12. Budd 4.5-in. rocket (top), Cenco 3.25-in. rocket (bottom), both with T-5 fuzes.

photographic observations of rockets in flight,
and of their limitations, are given in reference
7. Where possible, duplication of the informa-
tion given there has been avoided in the present
chapter on the testing of rocket fuzes.

Rockets and Launchers

In the chronological order in which the
rockets became available, the rockets used in
fuze testing and some of their characteristics
are given in Table 2. Some of the rockets were
fired with a variety of propellent charges, but
in general the table gives characteristics for

only those combinations most frequently used
and is intended to give a general picture of the
variety of test vehicles available. The availa-
bility of rockets of rather widely different char-
acteristics made possible a greater variety of
test conditions. Special values of acceleration or
maximum velocity could be provided as needed
to test fuze components under conditions more
severe than expected in service. Also airspeeds
could be obtained when firing from a ground
launcher equal to those expected in firing from
airplanes.

The ballistic characteristics are functions of
temperature and other conditions. The values
are representative of firings in summer.

Table 2. Rocket characteristics.

Fuzed rocket

Flight at 45°
quadrant

Diameter

Head

Unfuzed rocket

Burning

Max

elevation

motor

head

length

weight

distance

Time

vel.

Accel.

Range

Time

Rocket

(in.)

(lb)

(in.)

(lb)

(ft)

(sec)

(fps)

(g)

(ft)

(sec)

Cenco

3.25

.12

675

175

10,000

M-9, etc.

4.5

.20

925

1£0

12,500

3.25

3.5

600

1.0

1,500

18,000

3.25

3.5

500

1.0

1,150

16,000

3.25

5.0

450

1.0

825

13,000

HVAR

5.0

123

800

1.0

1,375

29,000

T-83

4.5

250

0.4

950

18,000

SECRET

THE FIELD TESTING OF ROCKET FUZES

327

As explained in Chapter 1, the development
of the T-5 rocket fuze for the M-8 rocket was
carried out concurrently with the development
of the rocket. This meant that no field testing
could be done until the rockets were available
unless some interim method could be devised.
To this end a simple inexpensive rocket was
designed and manufactured solely for the pur-
pose of carrying out experimental tests of the
fuze. This rocket was designed to give essen-
tially the same acceleration characteristics as

Figure 13. Four-rail rocket launcher, Fort
Fisher.

the M-8 rocket was expected to have. It had a
314-in. body. The first models were built in the
National Bureau of Standards shop and later
the Central Scientific Company manufactured
sufficient quantities for the field tests. It was
commonly referred to as the Cenco rocket. De-
tails of the construction are not included here
but may be obtained from reference 40. Some
testing was done with British rockets but their
acceleration characteristics were so different
from those of the M-8 rockets that their useful-
ness in testing T-5 fuzes was very limited. A
Cenco rocket with a T-5 fuze is shown in Fig-
ure 12. A multigrain propellent charge of sol-
vent-extruded double-base powder having a
total weight of about 2.65 lb was normally used
in this rocket.

The Cenco rockets were usually fired from
two-rail or from four-rail launchers which were
fabricated from iron pipes and other iron pieces

in the shops of the National Bureau of Stand-
ards and of the Aberdeen Proving Ground. Fig-
ure 13 is a photograph of a four-rail launcher
at Fort Fisher mounted on a truck chassis.

Two-rail launchers were more simple in con-
struction, but, since they were merely a pair of
rails upon which the rocket slid when fired, they
provided no restraint for the upper half of the
rocket, which therefore could, and sometimes
did, lift up from the rails at the forward end
and leave the launcher at a higher angle of ele-
vation than that of the launcher.

Figure 14 is a photograph of the forward end
of a two-rail launcher, constructed, in this case,
of two lengths of railroad rails. This launcher
was constructed at Fort Fisher at a time when
rocket motor blowups on the launcher were so
frequent, sometimes every other round, that
launchers constructed of pipes, which are usu-
ally destroyed in such a motor failure, were
impracticable. This railroad rail launcher suf-
fered many motor failures without suffering
sufficient damage to prevent its continued use.

The M-8 rockets became available in quanti-
ties sufficient for the field testing of fuzes about
the time that laboratory development of the

Figure 14. Two-rail launcher and target range,
Fort Fisher.

fuze was completed. They were then used for
testing production models of the fuzes, and the
Cenco test rocket was accordingly abandoned.

The M-8 series designation indicated HE
loading, and these rockets had inert-loaded
counterparts which bore a series of M-9 desig-
nations. There was no unanimity in the proper
designation of such rockets having neither inert
nor HE loading, and such empty rockets were

328

FIELD TESTING OF PROXIMITY FUZES

variously designated as M-8, M-9, empty M-8,
empty M-9, etc. There were other variations in
rockets, such as manufacture, fin design, and
powder trap design, which had a bearing on
fuze performance (see Chapters 5 and 9).

All these Army rockets had folding fins, as
shown in Figures 12 and 15. Toward the end of

Figure 15. Army 4.5-in. rocket and pipe
launcher. Blossom Point.

World War II, the rockets of this type were
supplied with attachable fixed fins and were
then designated as T-74 with no differentiation
with regard to loading.

The 4.5-in. rockets with folding fins could be
fired from ordinary iron pipes of suitable inside
diameters (the inside diameters of most of the
launching tubes used lay between 4.6 and 4.7
in.) such as the one of which the breech end is
shown in Figure 15. Consequently the construc-
tion and mounting of launchers of any desired
length for these rockets was a comparatively
simple matter.

At the time when tests of fuzes on Navy air-
craft rocket [AR] and high-velocity aircraft
rocket [HVAR] started, standard Navy launch-
ers, such as the one shown in Figure 16, were
already available and were very suitable for
proving ground use.

Additional details of rocket launchers and
firing procedures will be found, in the form of
informal notes, in reference 38.

Loading and Assembly of
Fuzes and Rockets

The assembly of rocket fuzes was very simi-
lar to that of bomb and mortar fuzes as de-
scribed in Sections 8.2.4 and 8.4.3 and does not
need to be given in detail. Further details on
loading, assembly, and storage procedures are
given in reference 32.

Except for the T-5 and T-6 fuzes for which
very satisfactory black powder spotting charges
were provided as a standard component by
Picatinny Arsenal, the spotting charges were
the same cartridges of a mixture of potassium
permanganate and magnesium powders as were
used in bomb and mortar shell fuzes. To mini-
mize fragmentation of the rocket head, which
was a hazard when early malfunctions occurred,
the rocket heads were often provided with four
1-in. holes to facilitate the emission of flame and
to reduce the pressure inside the head developed

Figure 16. HVAR rocket on Navy launcher,
Blossom Point.

by the explosion of the tetryl and spotting
charge. The 3.5-in. AR heads which appear in
Figure 17 are so drilled. Similar holes may be
seen in the photograph of the Cenco rocket in
Figure 12.

The propellant for many rounds of Cenco and

THE FIELD TESTING OF ROCKET FUZES

329

Figure 17. Groups of fuzed projectiles, Blos-
som Point.

Figure 18. Fuzed AR and HVAR rockets; fir-
ing tower, Blossom Point.

Army rockets was installed at the proving
grounds. Since these rockets are obsolete, refer-
ence is made to the notes given in reference
32 for details of loading and assembly. The
Navy rockets were fired as received, except that
particular care was taken to insure tightness of
the tail assembly, as described in the notes re-
ferred to above. Figure 18 shows a group of
fuzed AR and HVAR rockets at Blossom Point
ready to be taken to the firing point.

manner to determine the proportions of proper
and improper functions. The number of rounds
fired, when feasible, and the significance of the
results were determined by standard statistical
considerations (see Chapter 9).4>36’30 The most
commonly used firing methods were (1) at
a mock-aircraft target from a stationary
launcher, (2) high-angle firing tests against
ground or water targets, and (3) firing from
an airplane in flight.

Classification of Field Tests

The types or classes of information desired
and obtained from proving ground tests on
rocket fuzes are listed below. Reference is made
to the sections in which the methods used to
obtain particular types of data are described.

1. Fuze quality (Section 8.3.5) .

2. Fuze sensitivity and burst surface (Sec-
tion 8.3.6).

3. Fuze arming distance (Section 8.3.7) .

4. Causes of fuze malfunctions and effective-
ness of remedies (Section 8.3.8).

5. Exterior ballistics of rockets as affecting
fuze design and performance (Section 8.3.9).

Tests of Fuze Quality
General Procedure

Tests of fuze quality consisted in firing a
sufficient number of rounds in some particular

Target Tests (Horizontal Firing)

In most target tests efforts were made to
simulate conditions of air-to-air tactics. The
rockets were fired almost horizontally at a
mock-airplane target from a launcher mounted
on a substantial tower.

Since the fuzes are sensitive to approach to,
or departure from, a reflecting surface it was
necessary to select the firing conditions so that
ground reflection would not interfere with or
mask the aircraft target signals. If the fuzes
were fired horizontally over completely level and
electrically homogeneous terrain, they would
receive no firing pulses because there would be
no relative vertical component of velocity be-
tween fuze and terrain. While this situation
cannot be realized physically, it was found pos-
sible to choose terrain and height and elevation
of trajectory such that ground reflection signals
would be negligible.3 Such signals may arise
not only from relative velocity between ground

330

FIELD TESTING OF PROXIMITY FUZES

and target but also from sudden variations of
the reflection coefficient of the surface under
the trajectory. (Changes in reflection coefficient

constructed of well-bonded chicken wire with
wood supports. Figure 19 is a general view of
the range. Figure 20 is a close-up view of the
target. Notes on the method of suspending this
target are given in reference 33.

A view of the similar range at Blossom Point
is shown in Figure 8, a diagram of this range
in Figure 21, and a photograph of the 60-ft
tower may be seen in Figure 18. The dimensions

Figure 19. Horizontal target range, Fort
Fisher.

would change the load on the oscillator in the
fuze.) It was found that, with a target mounted
50 ft or more above ground and a launcher
about 1,000 ft away and at approximately the
same elevation, ground reflection signals due
to relative velocity were negligible. The eleva-
tion of the launcher was adjusted so that the
peak of trajectory was between the arming
point and the target. This arrangement'’ per-
mitted very rapid firing under conditions which
were essentially reproducible at any time and
had other advantages which are discussed in
Section 8.3.6.

Figure 20. Mock-plane target, Fort Fisher.

The first range of this type was set up at
Fort Fisher. The launcher was about 40 ft above
ground, the target about 60 ft above ground
and 1,000 ft from the launcher. The target was

T Target

S Side camera station
FT Firing tower

C Firing point camera, 50 ft above ground and directly
under projector

Target poles enclosed an area 100x125 ft

Figure 21. Diagram of target range at Blossom
Point.

of the target, which was usually suspended 70
to 75 ft above ground, are given in Figure 22.

As indicated in Figure 21, camera and ob-
serving stations were located in the towers di-
rectly below the launcher (with armor plate
beneath the launcher) and at side stations
located on lines normal to the trajectory at the
target. A view of the side observation station
at Fort Fisher is shown in Figure 23.

While visual observations were sufficient for
tests of fuze quality, moving pictures of the
spotting charge burst were usually taken in
addition. These supplied the more accurate
data needed for determining fuze sensitivity

SECRET -

THE FIELD TESTING OF ROCKET FUZES

331

and burst surface (see Section 8.3.6). Notes on
observational procedure will be found in
reference 34.

High-Angle Firing

When information on the performance of a
fuze under more severe conditions was desired,
the fuzed rockets were fired from the ground
to function upon approach to water after a long
flight. Such a test was more severe, since there
was more opportunity for malfunctions to occur
caused by generator failure, rocket vibration,
and fuze microphonics. The angle of elevation

Figure 22. Diagram of mock-plane target
(Blossom Point).

was raised sufficiently so that at arming the
fuze would be sufficiently far from the ground
not to be caused to function by the radio waves
reflected from the earth. Much firing was done
at elevations of 65 or 70 degrees.

In high-angle firing, visual and photographic
observations were made from stations suitably
located. The locations of stations often used at
Blossom Point are shown in Figure 11.

Figure 23. Side observation station, Fort
Fisher.

Firing from Airplanes

While the maximum airspeeds to be expected
in firing HE-loaded rockets from airplanes
could be attained in launchings from ground
locations by the use of lighter rockets, other
conditions obtaining in actual service use, such
as the initial airspeed and vibration of an air-
borne launcher, could not be readily dupli-
cated. Consequently, from time to time tests
were made in which fuzed rockets were fired
from airplanes, usually for function upon ap-
proach to land or water. Reference 20 gives the
results of a typical test of this type at Dahlgren.
Reference 22 gives the results of a test at
Aberdeen.

Carrier Indications of Fuze Performance

As an aid in determining the causes of fuze
malfunctions, for nearly every round fired the
presence or absence of fuze carrier was deter-
mined, and for nearly all rounds of generator-
powered fuzes, phonograms or cathode-ray
oscillograms of carrier modulation were ob-
tained. The circuits used are described in de-
tail in Section 8.4.3. Figure 24 is a photograph
of the radio receiving and recording equipment
used at Blossom Point.

Carrier modulation records gave evidence of
microphonics, generator bearing seizure, and
generator frequency. To determine generator
frequency the approximate values to be ex-
pected had to be known, since filament voltage
ripple produced fundamental modulation fre-

SECRET

332

FIELD TESTING OF PROXIMITY FUZES

quency and plate voltage ripple twice funda-
mental, and either might predominate. To
obtain unambiguous generator speed values, re-
porter units were used on rockets as discussed
for bomb tests. Analysis of the records was as
discussed in Section 8.2.12.

Figure 24. Radio receiving and recording equip-
ment, Blossom Point.

8'3'6 Determination of Fuze Sensitivity
and Burst Surface

General Considerations

In order to correlate laboratory data on the
electric characteristics of fuzes with expected
Service characteristics, many tests were per-
formed to determine the probability and locus
of function of fuzes passing or approaching
specified targets at measured speeds and dis-
tances. Such data were necessary in order to
assess the effectiveness of a given fuze for a
given tactical application. An example of such
an assessment is presented by a series of re-
ports of the Applied Mathematics Panel (refer-
ences 19, 20, and 21 of Chapter 1).

Water-Approach Tests

One of the simplest methods, and one often
used, to determine the sensitivity character-
istics of a given type of fuze is to measure the
height of function upon approach to water. The
reflectivity of the target (water) is essentially
constant at a given location, and the variation

in sensitivity with aspect can be investigated
by varying the angle of approach, giving due
allowance to the varying vertical component of
the approach velocity. In such tests, assuming
the ballistics of the fuzed rocket are known
(cf. Section 8.3.9), the proving ground is
called upon to provide the photographic records
from which the height of function over water
can be determined.

The determination of height of burst was
often combined with the obtaining of other
types of information, such as afterburning
characteristics (cf. Section 8.3.8) and con-
sequently was sometimes done in daylight,
sometimes at night.

The method of obtaining function heights in
daylight was the same as that for obtaining
function, heights of bomb fuzes, except that
aiming circles were used to obtain triangula-
tion data. Two or more observation stations on
surveyed base lines as long as were necessary
or practicable were used.

At each station were an aiming circle opera-
tor and a 16-mm moving picture camera opera-
tor. Each station was provided with telephone
or radio communication with all other points
of operation.

The distance of the point of function from
each camera was obtained by use of large plot-
ting boards. Then since the effective focal
lengths of the lenses used had been determined
by photographing scales of known lengths at
known distances, as described in Section 8.2.9,
the measured heights of function on the films
could be translated into actual heights. The
measured distance on the film was the distance
between the splash at impact and the first
visible flash of the spotting charge. Since the
distance from camera to function was always
large in comparison to the height of the- camera
above the water surface, the fact that the splash
was not directly beneath the point of function
produced no significant error.

At night the splash did not appear, and the
method of determining the height of function
consequently was somewhat different. Still
cameras, instead of moving picture cameras,
were used to photograph the flash of the spot-
ting charge and lights whose positions and
heights above water were accurately known.

SECRE1

THE FIELD TESTING OF ROCKET FUZES

333

The reference lights were either navigation
lights or temporary lights placed at intervals
on a line crossing the area in which impacts
were expected.

To reduce the number of films used and de-
veloped in such tests, special cameras were de-
signed and constructed. These cameras carried
8xl0-in. films in holders which could be racked
down across a horizontal focal plane slit. As
many as 15 rounds could be photographed on
one film with such a camera.

The position of function was determined by
triangulation on a range plotting board, the
azimuth of the flash at each station being
determinable from its position on the film with
respect to the reference lights. By trigono-
metrical methods, the apparent height of flash
with respect to the reference lights could then
be translated into actual height of function
above the surface of the water. A value for the
height of function was obtained for each
camera and the average agreement in these
values was of the order of y20 of 1 per cent of
the distance from camera to function.

Target Tests

While water-approach tests gave data on the
sensitivity of the fuze with respect to a large
horizontal reflecting surface, target tests were
required to determine whether the position of
the function with respect to an airplane was
favorable for producing effective damage. When
sufficient rounds were fired, the positions of
burst obtained defined a surface, or rather sur-
faces (because of the radiation lobes of the
fuze), on which function was most probable as
a fuze passed the target.

Before horizontal ranges came into use, such
tests were performed, mainly at Fort Fisher,
by firing fuzed rockets at targets supported
from balloons, using the mobile launcher shown
in Figure 13. The burst of the spotting charge
was photographed from two stations, one of
which was approximately directly beneath the
target, the other off to the side. A camera was
also placed near the firing point at times.
Especially built fixed-focus cameras using
8xl0-in. plates and equipped with azimuth and
elevation scales were used in these tests. The
target was generally an array of crossed dipoles,

the array being 40 ft long. Figure 25 shows a
burst on such a target. The balloon used for
supporting this type of target is shown in Fig-
ure 26 in its hangar at Fort Fisher. The known
length of the target was a useful parameter in
determining the position of function with re-
spect to the target.

Because the target position was continually
changing, and dispersion was large because the
launcher had to be short to be mobile, these
tests were time consuming and wasteful of
ammunition. Many rockets passed the target at
distances too large to permit the fuze to func-
tion. With the installation of the ranges for
horizontal firing described in Section 8.3.4,
balloon-supported targets were abandoned and
the testing greatly expedited. Since the target
was fixed, the launcher had to be only slightly
adjustable and consequently could be made
sufficiently long (30 or 40 ft) to reduce ma-
terially the dispersion of fast-burning rockets.
Considerable control of the position of the
trajectory with respect to the target was then
possible. Moreover, the aspect of the target,
which was constructed to have approximately
the reflectivity pattern of an airplane, could
be readily changed as desired. Still another
considerable advantage was the fact that the
camera positions were fixed and could be
located in such positions (directly beneath the
launcher and on a line normal to the centerline
of the range at the target) as to reduce to a
minimum the operations required to determine
the position of function.

8 3 7 Determination of Arming Distance
or Time

All proximity fuzes were designed to arm at
a distance determined by various considerations
of safety and tactical use. Arming tests in the
field were performed to determine the reliability
of arming mechanisms under standard and
marginal conditions and to obtain data from
which the statistical distribution of arming
times, or distances, could be computed.

The arming arrangement was mechanical or
electric (RC arming) or a combination of the
two. With any of these arrangements, the

334

FIELD TESTING OF PROXIMITY FUZES

determination of the time of completion of the
mechanical arming process was comparatively
simple and direct. The time to completion of
RC arming was more difficult to determine.

Photographic Method
(Mechanical Arming)

In the most direct, and most frequently used,
method for investigations of mechanical arm-

method could be used only with generator-
powered fuzes. It involved modifying the fuze
so that at the completion of mechanical arming
a recognizable change in the modulation of the
carrier would occur. The fuze wiring was
changed so that the detonator rotor contacts
were in the filter and rectifier circuit. At arm-
ing, then, a change in amplitude or frequency
of modulation or both occurred and was re-

Figure 25. Rocket with smoke tracer and function
Fort Fisher.

ing, the fuze was modified to function upon com-
pletion of mechanical arming, mounted on a
rocket and fired in the usual manner. The time
at which the fuze functioned was then measured
by stopwatch, or the time and locus of func-
tion were determined photographically. Refer-
ence 14 includes a description in detail of a
method for obtaining indications of a number
of arming functions on one film or plate and for
interpreting the results.

Carrier Indication of Mechanical Arming

A radio method of measuring the time to
completion of mechanical arming was used
where feasible, since it had the advantage of
not destroying the fuze at arming and so allow-
ing the determination of arming time to be
combined with other types of testing. This

on array of crossed dipoles suspended from balloon,

corded on an oscillograph or sound-recording
equipment connected to the output of a radio
receiver. The time of launching was obtained
on the same record by mounting a switch in a
1,000-c circuit on the launcher in such a posi-
tion as to be opened or closed by a fin or other
part of the rocket as the rocket started to move
forward.

Since the time to arming was often too short
to allow the carrier to be tuned in with a re-
ceiver of normal selectivity, broad-band re-
ceivers (made by Zenith) were often used.
These receivers have a flat frequency response
over a range of ±3 me. Their main drawback
is the absence of an r-f stage with consequent
possibility of direct interference through the
i-f stages. In addition, they are inherently less
sensitive than receivers of greater selectivity

SECRET

THE FIELD TESTING OF ROCKET FUZES

335

and consequently could not always be used
when desired.

RC Arming Measurements

A possible method of determining the time or
distance to completion of RC arming is to vary
the distance between launcher and target and
obtain the distribution of duds and proper func-
tions as a function of this distance. Some ex-
periments of this type were performed by firing
from airplanes to ground as described in this
section. Otherwise there was no satisfactory

Figure 26. Balloon in hangar, Fort Fisher.

method of varying the distance between the
launcher and a physical target. Instead, in hori-
zontal firing, a portable sweep-frequency trans-
mitter was used to supply a triggering pulse to
fuzes in flight at various positions along the
trajectory.

Because of rocket dispersion, the power of
the transmitter had to be greater than would
have been necessary if the distance of passage
had been constant. Consequently, if the trans-
mitter had been left in continuous operation,
the position at which the fuze first received a
signal of firing intensity would have been in-
definite and could have been up to 200 ft short
of the point of passage above the transmitter.
For this reason, a time switch, consisting of a
thyratron and associated RC circuit initiated
by a rocket-operated switch on the launcher,
was used to key the transmitter at the time
when the rocket was directly over the trans-
mitter. The interval of time between launching

and operation of the transmitter was measured
automatically by means of a time clock.

It was not possible to apply continuous
signals or a series of signals because if a pulse
of sufficient magnitude to fire the thyratron
were received by the fuze before arming was
complete, the detonator-firing capacitor would
“dump” the charge and the arming cycle would
start over (see Section 3.3.6). Thus the pulsing
method of measuring arming times gave only
a “yes” or “no” indication on each round fired.
Large numbers of rounds had to be fired to
obtain reliable arming time data.

This arrangement was used in arming tests
of the T-30 fuze on HVAR and AR rockets. Be-
cause of the cessation of hostilities these tests
were not so extensive as originally planned. A
progress report on the results is given in the
Bibliography.38

Function, No-Function Tests at
Various Slant Ranges

In order to test arming characteristics under
Service conditions, fuzed rockets, inert or HE-
loaded, were fired from planes at various slant
ranges. By determining the ranges of firing
photographically, the dividing line between
duds and proper functions could be bracketed
and the arming distance determined to a degree
of certainty dependent upon the number of
rounds fired. This method was applicable to
the testing of fuzes of any type with any type
of arming.

8'3'8 External Causes of Fuze Malfunctions
Introduction

Throughout the development of proximity
fuzes for rockets, much field testing was di-
rected toward investigating causes, external to
the fuze, of malfunctions and the effectiveness
of remedial measures devised to minimize such
effects. Propellant afterburning and instability
of rocket parts were particularly troublesome.
Other factors studied were the effect of temper-
ature upon setback, upon which arming de-
pended, the effect of raindrops upon fuze per-
formance, and the effect of rocket spin upon
the arming of the T-5 and T-6 fuzes.

336

FIELD TESTING OF PROXIMITY FUZES

Investigations of Afterburning

Afterburning may be defined as the delayed
burning of remnants of propellant or other
combustible material which, for one reason or
another, remain in the combustion chamber
after main burning has ceased. (See Section
9.2.2.) The ionization produced by afterburning
may cause malfunctioning of the fuze. (See
Section 2.13.) Numerous field tests were de-
voted to studying the effect with many fuze-
rocket-propellant combinations.24’ 38

Much of the firing in these investigations was
done at high angles and at night, in order that
visual or photographic observations of after-
burning might be correlated with observations
of the locations of fuze functions.

A considerable accumulation of data was
usually required before the incidence of mal-
functions caused by afterburning could be
differentiated from the incidence of malfunc-
tions resulting from other causes. If afterburn-
ing was observed to occur only in the first part
of the trajectory, the observed distribution of
malfunctions in the rest of the trajectory could
be extrapolated back into the afterburning
region and the number of malfunctions due to
other causes in this region subtracted from the
total in this region to give a residue, the major
portion of the total in such a case, attributed
to afterburning. In other cases the analysis
and interpretation were less straightforward
and consequently less satisfactory.

As was to be expected from the frequency
selectivity characteristics of the fuzes, intense
afterburning was not necessarily accompanied
by a high incidence of malfunctions. This was
supported by static tests in which the pulses
produced in the output circuit of the fuze by
afterburning were recorded as cathode-ray
oscillograms and correlated with simultaneously
obtained moving pictures of the actual phe-
nomena occurring and with performance in the
field.18 In general, it was found that a triggering
flame was always a visible flame, but that all
flames did not necessarily give rise to transients
capable of triggering the fuze.

Figure 27, a typical set of photographs from
night-firing tests, illustrates the great varia-
tions in afterburning phenomena encountered.

The M-9 rockets were used in these rounds. The
appearance of main burning also differs
markedly with different propellants, as illus-
trated in Figure 28.

Effects of Rocket Structure

At one time or another almost every possible
source of mechanical or electric instability in
rockets was suspected and investigated as a
cause of fuze malfunctioning. The method of
investigation was the obvious one of statistically
analyzing the incidence of malfunctions before
and after making a modification in the rocket
designed to eliminate the suspected source of
triggering pulses. As examples, studies were
made of the effect of electrically bonding the
joint between head and motor, of the effect of
electrically bonding the joint between folding
fins and motor (leaving the fins still movable),
of the effect of welding the fins rigidly in the
open position, of the effect of making the trap
structure more rigid, and of the effect of rocket
spin (see Chapter 9 and reference 21) upon fuze
performance.

The studies of fin structure led to a recogni-
tion of the necessity of inspecting the locking
action of folding fins and to the design of a
crimping tool which was used to make the lock-
ing action of individual fins more positive where
necessary. This tool became a standard serv-
ice item and was provided for use in combat
areas.

Effect of Propellant Temperature
upon Arming

Since the arming of rocket fuzes depends
upon acceleration, and acceleration is affected
by propellant temperature, tests were made to
determine whether the arming mechanism
would operate properly at the extreme service
temperatures of the rockets. The fuzes were
either arranged to function on arming or the
incidence of duds with normal fuzes at the ex-
treme temperatures was investigated. In these
tests the rocket motors were first brought to
the desired temperature in thermostatically
controlled chambers. They were removed and
fired quickly before the powder temperature
changed appreciably. A typical test of this type
is described in reference 16.

SECRET

THE FIELD TESTING OF ROCKET FUZES

337

Effect of Raindrops upon
Fuze Performance

Tests on the performance of fuzes, with and
without plastic shields, in rain and in clear
weather revealed that impact with raindrops

drop size and concentration at the time of firing,
a method of obtaining permanent records of
these quantities was developed.26 Outing flannel
was impregnated with a mixture containing
methyl violet. When a square of this material

REGULAR POWDER
LEAST AFTER - BURNING
EARLY AT 2.1 SEC
(L-I7S3)

REGULAR POWDER
AVERAGE AFTER - BURNING
EARLY AT 2-4 SEC
(L- 1729)

REGULAR POWDER
MOST AFTER - BURNING
EARLY AT 1.0 SEC
(L- 1735)

DRY POWDER
AVERAGE AFTER- BURNING
EARLY AT 3.4 SEC
(L- 1755)

PUFF AND AFTER -BURNING
PICTURES , TAKEN WITH
CAMERA PLACED ABOUT 30
FEET TO SIDE OF PRO-
JECTOR AND POINTED TO
COVER FIRST FEW SECONDS
OF FLIGHT. INITIAL BLAST
PARTIALLY OR COMPLETELY
EXCLUDED

WET POWDER
AVERAGE AFTER -BURNING
EARLY AT 1.4 SEC
(L- 1736)

A-41 POWDER A-41 POWDER

NO EARLY FUNCTION EARLY AT 2.2 SEC

(L- 1757) (L- 1764)

PURGE PELLETS
NO EARLY FUNCTION
CL- 174 4)

Figure 27. Afterburning with various propellants, M-9 rocket, Fort Fisher.

could cause malfunctions the incidence of which
could be reduced by the installation of plastic
caps.

In order to obtain quantitative data on rain-

was exposed to rain for a measured length of
time, a purple spot was formed for each drop
striking the cloth. The diameter of the spot
was found to be about 85 per cent of the diam-

338

FIELD TESTING OF PROXIMITY FUZES

scale: i i i i —

O 10 30

RD B 11131

_! ! I

50 75 100 FEET

RD B 11137

MJX PROPELLANT

Figure 28. Main burning with two propellants, T-83 rocket, Blossom Point.

THE FIELD TESTING OF ROCKET FUZES

339

eter of the drop. From such a record, the
number of drops striking unit area in unit time
and the diameters of the drops could be deter-
mined.

Sympathetic Functioning in Rapid Firing

Tests of fuzed HE-loaded rockets, launched
in quick succession, were made to determine
whether the fuzes would function sympa-
thetically, that is, whether the ionization or
fragments produced by a burst would cause the
fuzes on adjacent or succeeding rockets to func-
tion also. One fuze in each group of rockets
fired in rapid succession was modified to func-
tion at a predetermined time. A rotary, multiple-
contact firing switch driven by a spring motor
was provided to fire the rockets with a desired
interval of time, about y10 sec, between suc-
cessive rockets. Moving pictures and visual
observations were made to determine the time
and location of bursts.11

839 Investigations of the Exterior
Ballistics of Rockets

Introduction

Since the weights and ogives of the rockets
as used were seldom standard, the ballistic data
required in the course of the development of
rocket fuzes were usually obtained at the prov-
ing grounds, often simultaneously with tests of
fuze performance. Quantities measured were
velocity, acceleration, range, angle of terminal
approach, rate of spin, and yaw.

The instruments used in these measurements
included standard 16-mm moving picture
cameras, a Western Electric 16-mm high-speed
camera, Hickman 8-mm high-speed cameras,
ribbon-frame cameras, still cameras with
rotating shutters, ballistic coils, and a photo-
electric-radio yaw telemeter. The characteristics
of the photographic instruments are given in
Chapter 13 of reference 13. Chapter 4 of the
same publication describes the mathematical
procedures used in trajectory calculations (see
also Section 8.4.4) .19- 35> 45> 46

Velocity, Acceleration, and Range

Velocities were determined photographically
if values of velocity in a relatively short portion

of the trajectory (several hundred feet or less)
were desired; the data were obtained in day-
light, using high-speed cameras or ribbon-
frame cameras or both. If velocities throughout
several thousand feet of trajectory were de-
sired, the rockets were equipped with flame
tracers and fired at night. Still 8xl0-in. cameras
with slotted disk shutters driven by synchro-
nous motors were used to obtain the position of
the rockets at known intervals of time. Refer-
ence lights were used to establish a scale of
distance from the launcher.

None of these methods was capable of giv-
ing velocity curves from which reliable accelera-
tion curves could be obtained but did suffice to
give average values of acceleration. Atmos-
pheric refraction at night was a particularly
troublesome source of error.

In horizontal firing from a tower, a timing
disk driven by a synchronous motor and a small
mirror through which the launcher could be
photographed were placed in the field of the
moving picture camera at the side station. From
the films so obtained, the average velocity from
launcher to target and the velocity at the time
of passing the target could be determined
whenever desired.

When the ribbon-frame cameras were new, it
was found that the synchronous motor drive
could be depended upon to give exposures at
twice line frequency at voltages greater than
90 v, but after repeated use it was found that
the motor drive could no longer be depended
upon. The cameras were then equipped with
neon-bulb timing devices in which the light
from the bulb, which flashes at twice line fre-
quency, was led directly to one edge of the film
through a Lucite rod tapered to produce a
narrow trace upon the film. At the same time
it was determined that velocities obtained
photographically and by means of ballistic coils
and a cathode-ray chronograph arrangement
were in agreement.25

Range determinations were made by triangu-
lation from two or more observing stations as
described in Section 8.3.6.

Angle of Approach

More reliance was placed upon angles of ap-
proach obtained from trajectory calculations

340

FIELD TESTING OF PROXIMITY FUZES

based upon determinations of maximum veloc-
ities and ranges at various angles of elevation
than upon angles obtained by photographic
means. Attempts were made to obtain the angle
of approach directly by equipping rockets with
tracers and using cameras placed approxi-
mately on a line normal to the line of fire at
the point of impact, but it was found that
atmospheric refraction over water at night
introduced errors of such a magnitude that the
results could not be trusted.

Determination of Rate of Spin

Since the rockets used were fin stabilized,
spin, when it occurred, was usually accidental
and had a speed of not more than several
hundred revolutions per minute. Consequently
the rate of spin was easily measurable. The
rockets used for this purpose were painted half
white and half black and photographed in flight
with a ribbon-frame camera located off to the
side.17

By this technique it was found that accidental
tilting or bending of the fins of M-9 rockets
would produce spin of the rocket. This spin
caused malfunction of the arming switch of T-5
and T-6 fuzes (see Section 9.2. 2). 21

Measurement of Yaw

A few measurements of yaw were made. The
frequency of yaw of rockets equipped with
smoke tracers and fired from a plane was deter-
mined from ordinary 16-mm moving picture
film.8 * For one round the frequency and ampli-
tude of yaw were measured during flight by
means of a photoelectric-radio telemeter, using
the sun as a localized source of light.10

8 4 THE FIELD TESTING OF MORTAR
SHELL FUZES

Introduction

All mortar shell fuze testing under the
auspices of Division 4, NDRC, was performed
at Blossom Point (see Section 8.3.1) and at
the Clinton Field Station of the University of
Iowa. Figure 29 is a graph showing accumula-

tive totals of rounds fired at the two proving
grounds.

Figure 29. Accumulative number of mortar
shells fired.

The Clinton Field Station was located along
the Mississippi River about 2 miles north of
Clinton, Iowa. Figure 30 is a map of the prov-
ing ground and Figure 31 a photograph of some
of the buildings in the central area. Figure 32
is a photograph of the view down range from
Tower No. 1 and Figure 33 is a view of Tower
No. 3, which included a fragment-proof shelter.

The testing procedures used at Blossom Point
and at Clinton were very similar. The Clinton
Field Station was designed exclusively for the
purpose of testing mortar shell fuzes. Conse-
quently, the practices obtaining at that proving
ground have been considered particularly
pertinent in preparing this section of Chapter 8.

One essential difference between the testing
of mortar fuzes and other proximity fuzes was
in the provisions for taking ballistic data.
Range and velocity measurements were taken
on most rounds of mortar tests in order to
provide data on the effect of the fuze on the
flight of the missile.

SECRET

THE FIELD TESTING OF MORTAR SHELL FUZES

341

Point on line

-0-

Telephone pole

-©■

Power pole

Base line target

Fence

Gun station

Storage

Loading shack

Radio shack and tower

Office

Machine shop

Equipment room
Head

Temporary storage
High explosive storage
Guard quarters
Guard house at front gate
Detonator storage

KMNO4 and magnesium puff storage
Board walk from gun station to No. 10
Added power lines
Added fence

Figure 30. Map of Clinton Proving Ground.

342

FIELD TESTING OF PROXIMITY FUZES

8.4.2 Personnel and Equipment

Figure 34 shows the distribution of personnel
and equipment for a typical test of fuze quality.
In addition to the 14 men for whom duties are
listed in Figure 34, three were used in loading
operations and one for carrying ammunition.

Figure 31. View of Clinton Field Station.

The work of developing and reading films, com-
puting and analyzing data, writing reports and
handling business details was done in an office
building in Clinton, 3 miles from the field sta-
tion. The personnel in this office included four

Figure 32. View of towers and gun position at
Clinton Field Station from T± tower.

persons to assemble data and perform calcula-
tions, one operator for the film developer, two
film readers, two secretaries and two men to
analyze data and write reports. A report could

be completed in 4 hours after the raw data
were received from the field.

8-4 3 Operating Procedures

Coordination of Firing

It was necessary to set up a definite routine
and to exert considerable effort in coordination
of the firing operation in order to carry out
smoothly a firing program of 100 or more
rounds per day. All men were familiarized with
the test program before going to their stations
and were kept supplied with pertinent informa-

Figure 33. View of Ts tower at Clinton Field
Station.

tion during the firing program. The firing oper-
ation began with the gunner placing the shell
in the release mechanism and asking for clear-
ance from the tower. The operator in charge
at Ti ascertained clearance and each of the sta-
tions informed the gunner of readiness. On the
informative count of ten, the gunner caused
the release mechanism to drop the shell down
the mortar barrel.

Knowing the approximate time of flight of
the rounds, it was a simple matter for the

THE FIELD TESTING OF MORTAR SHELL FUZES

343

operator in charge to inform everyone when
the unit was expected to function. At approxi-
mately 2 sec before this time, he gave the signal
“camera.” The cameras were started on this
signal and the aiming circle operators became
alert. Much film and time required for reload-
ing the cameras were saved in this manner. If
anyone observed an early function or a dud, he
immediately informed all operators over the
telephone. These methods relieved the opera-

sistors shown were separate resistors in order
to guard against excessive current in case of a
short circuit of one of them. Figures 38 and 39
show the fixture into which the rotor was in-
serted and the testing meter constructed at the
field. Identical testing devices were used for the
T-132 and T-172 fuzes except that the contacts
on the fixtures holding the rotors were different
because of the designs of the two rotors. Plac-
ing the interrupter plate (also called the lead

Bluff

Personnel for Each of the Towers

h‘ 14**

1. Camera Operator

2. * Aiming Cirole Operator

3. * Telephone Operator, who may

also operate a second aiming
circle*

Gun Position Personnel

1. Gun Operator — handles telephone

2. * Operator of Muzzle Velocity Machine

and B. C. Telescope

Personnel at Ti Tower

1. * Operator in Charge — handles telephone

2. Camera Operator

3. * Aiming Circle Operator

Personnel for Radio Building

1. Operator for Recordings

2. * Telephone Operator

3. Camera Operator to photograph oscilloscope

* These men also handled a stop watoh*

** T2 and*T4 did not operate simultaneously.

Equipment at Each of the Towers

12* 3* 14

1. Camera

2. Two Aiming Circles
3* Two Stop Watches

4. Telephone

5. Blackboard

Gun Position Equipment

1* Mortar with Associate Release Mechanism

2. Muzzle Velocity Measuring Ueohaniem

3. Stop Watoh

4. B. C. Telescope

5. Telephone

6. Blackboard and Clock to be photographed
from Tl

Equipment at Tl Tower

1. Camera
2* Aiming Circle

3. Two Stop Watches

4. Telephone

5. Speaker connected to Radio Receiver"
Equipment for Radio Building

1. Broadband Receiver

2. Oscillator

3. Mixer

4. Limiter

5. Recorder

6. Motor Driven Camera

7. Oscilloscope

8. Stop Watoh

9. Telephone

10. Headphones

11. Microphones

Figure 34. Personnel and equipment at observation towers, gun position, and radio building during
firing operations.

tors of undue tenseness and allowed them to be
alert at the proper time.

Loading Operations

The loading operation involved the assem-
bling of the components (Figure 35) into the
complete 81-mm shell with the VT fuze as
shown in Figure 36. A supply of the component
parts, except the VT fuze, was kept at the field
station in order ter allow much of the loading
operation to be carried out before the day of a
firing program. The loading of the detonator
into the rotor was the first operation. Figure
37 is a circuit diagram of the device used for
testing the rotor after it was loaded. The re-

or tetryl plate) in the fuze and screwing in the
booster cup completed the loading of the VT
fuze.

Shells as received contained a filler of bismuth
carbonate in paraffin wax. To facilitate observa-
tion and photography of the height of func-
tion, a cartridge containing a mixture of
potassium permanganate and magnesium metal
was used to provide a flash and smoke puff. A
cavity drilled in the shell filler provided a space
for this cartridge. The hole for the cartridge
was drilled with a 1-in. bit and reamed to a
diameter of 1% in. As the program closed, the
problem of drilling the filler of stearic acid and
plaster paris presented itself. Because this

SECRET

344

FIELD TESTING OF PROXIMITY FUZES

filler is so much harder than paraffin, it ap-
peared that drilling into it would require the
use of a power-operated setup built around a
drill press with an unusually long spindle
travel.

The M-56 fin was commonly used with the

Booster cup

KMNO4-MG puff cartridge

Booster pellet

Ignition cartridge

Interrupter plate

Fin (tail) for M-56 shell

(lead, tetryl)

Primer

Fuze

Spacing washer

Fuze rotor

Increment holder

Detonator

Smokeless powder incre-

M-43 shell

ments

Figure 35. Component parts of 81-mm mortar

shell with VT fuze.

M-43 shell body. To fit the fin to the body, it
was necessary to saw off the first two threads
from the fin and insert a spacing washer of
y16 in. thick aluminum between the shell and

Figure 36. 81-mm shell with VT fuze assem-

bled.

the fin. The shell was then completely assembled
with the smoke puff cartridge inserted and the
VT fuze screwed into place. After the weight
of the shell, the number of the round, and the
serial number of the fuze had been recorded,

the shell was ready for delivery to the gun
position.

Gun Position

The mortar was set up (Figure 40) and

500 400 400

Figure 37. Circuit diagram of rotor tester.

aimed in accordance with the instructions
given in the Basic Field Manual for this par-
ticular gun. Information as to elevation and
point of aim were furnished the gunner from

Figure 38. Fixture for holding loaded rotor.

the test request and weather data taken prior
to firing. The cage on which the solenoid coils
(used in muzzle velocity measurements) were
mounted was adjusted so that the axes of the

THE FIELD TESTING OF MORTAR SHELL FUZES

345

coils coincided with the axis of the gun. The
cage was adjusted to the same elevation as the
gun (Figure 41) and then shifted in a hori-
zonal plane until the axes coincided.

The person operating the electronic timing

and the range, the height of function was
readily obtained.

In addition to the apparatus mentioned above,

Figure 39. Rotor testing meter and jig.

device (Figure 42) for measuring muzzle
velocity also operated a battery command [BC]
telescope at the gun position. The BC telescope
was often used to get a quick check on heights

Figure 41. Solenoid coils used for measuring
muzzle velocity.

a clock and blackboard were located at the gun
position (Figure 32) and photographed from
T i between rounds. The information on the

Figure 40. Mortar shown in position for firing.

of function. It had a vertical mil scale on which
the angle between the smoke puff (or flash)
and the splash could be read. From these data

Figure 42. Electronic timing device for meas-
uring muzzle velocity.

blackboard was changed from round to round
by the same operator who handled the muzzle
velocity apparatus. A detailed description of

346

FIELD TESTING OF PROXIMITY FUZES

the muzzle velocity measuring mechanism is
given in reference 41.

Each shell, after inspection, was magnetized
so that it would actuate the muzzle velocity
apparatus by inserting it in the magnetizing
coil (Figure 43). The shell was then placed in

Figure 43. Shell being magnetized in magnet-
ization coil.

the special release mechanism (Figure 44) and
the mechanism placed on the mortar barrel
(Figure 45) . This device enabled the operator to
drop the shell down the mortar barrel by pulling
a string from behind the concrete barricade
(see Figures 40 and 42) . The gun operator was
then ready to ask for clearance from the T i
tower.

The gun operator was responsible for watch-
ing the early flight of the shell so as to note any
unusual behavior such as excessive yaw or
tumbling. The second operator recorded the
muzzle velocity, time of flight of the round, and
obtained the mil height of function in the BC
telescope. While the gun operator magnetized
the next round and placed it in position to be
fired, the second operator recorded the firing
point data and inserted the serial numbers for
the next round on the blackboard.

Ballistics Data

The purpose of this discussion is to explain
the method of measurement of the data, the
method of calculation of the point of function,
and the accuracy to be expected from the

Figure 44. Shell being placed in special release
mechanism.

methods and apparatus used. It was desirable
to know with a fair degree of accuracy the
point of function or point of impact of each
round. This information was necessary to
determine the range and deflection of the
round and the height of functioning of the unit.
The point of function was located in the event
of proper functions ; the point of impact in the
case of duds. The point of function will be used
here to include both cases.

Method of Taking Measurements. Figure 46
shows the location of the four towers, the gun
position, and the line of reference. Each of the
four towers had at least one aiming circle which
was used to measure the azimuth of the point
of function. The aiming circle is a device
similar to the transit : its calibration is in mils ;
its field is approximately 85 mils in radius, its
magnification 4 to 1 ; and its turning angle is
360 degrees or 6,400 mils (a mil on the aiming
circle is defined as 1/6,400 of a circle). The
scale of the aiming circle was in the field of
view, the azimuth of the point of function of

THE FIELD TESTING OF MORTAR SHELL FUZES

347

the round, with respect to the center of a scale
located in the field of view, was read at the
time of function. The reading thus obtained
was added to or subtracted from the angle of
the setting of the aiming circle, depending upon
whether the point fell to the right or left of the
center line. Since the area of view from the
aiming circles was small for points of func-
tion near the observation towers, it was at
times necessary to employ two overlapping aim-
ing circles in order to locate all of the points
of function.

The zero setting of the aiming circle at Ti
tower was along the line from the to the T2

Figure 45. Shell and release mechanism being
placed on mortar barrel.

tower, and the center of the aiming circle was
moved clockwise 328 mils to lie along the
reference line. The zero settings of the aiming
circles at T2, T3, and T4 towers were along the
line from these towers to Tx tower. The azimuth
of the point of function of the round was
measured from these zero settings in a clock-
wise direction.

Method of Computation. The following
symbols are used in explaining the method of
making computations and analyzing the errors
involved :

|3 = angle in mils between line of aim and
reference line.

Figure 46. Sketch of Clinton Field Station,
showing line of reference and position of obser-
vation towers and gun.

a = angle in mils measured at T1 tower
from the reference line to the point of
function. Both a and (3 were positive if
measured to the right of the reference
line and negative if measured to the
left.

T4 = 328 + a in mils, which is the angle be-
tween line of Ti and T2 towers and
point of function.

Ti — angle in mils at Tx tower measured
clockwise from T4 tower to point of
function ( i = 2, 3, 4).

AT. = indicates the amount of error of meas-
urement, in mils, of the angles T{ (i =
1, 2, 3, 4).

348

FIELD TESTING OF PROXIMITY FUZES

I SECRKT

|inwmtit|<ittluH| 1 1 1 i |iiti|im|nn|in i|n 1 1 1 1 1 n 1 1 H)-| 1 1 m | i|i ? m*t m 1 1 1 P 1 1 1 1 IM* 1 tf H H M H1 1 1 1 j 1 * 1 1 itntfiii|tiii|iiii|ini|iin|iiu|Hi|

4800 4400 4200 4000 3900 3800 3700 3650 3600 3550 3500 3450 3400 3550 3320 3300

THE FIELD TESTING OF MORTAR SHELL FUZES

349

TiF — the average of two calculations of the
distance in feet from tower to the
point of function, in case of discrepancy
giving greatest weight to the calcula-
tion using the angle measured at the
tower nearest to the point of function.
T1Fi = the distance in feet from T i tower to
point of function, calculated using
angles and Ti (i = 2, 3, 4).

R = range of the round in feet, i.e., the dis-
tance from gun position to point of
function.

T\Fi — the distance in feet from Tt tower to
point of function of the round, calcu-
lated using the angles measured at Ti
and Ti towers (i — 2, 3, 4) .

Knowing the angles T± and T { ( i = 2, 3, or 4)
and the distance from T4 to Tif the distance
from Ti to the point of function of the round
was calculated by use of the law of sines. Since
the angle T2 and T3, or angles T3 and T4 were
always obtained, there were two calculations
for each point of function to check against each
other. Similarly, T2Fly T3F1} T4Fi were calcu-
lated by use of the law of sines.

A nomograph was constructed to solve the
trigonometric relations between the above
terms. Figure 47 exhibits the nomograph used.
The final distance T4F was given as the average
of TiF2 and TXF3 or of TiF3 and 7\F4, giving
greatest weight to the value calculated with the
data from the tower nearest the point of func-
tion.

The range R of the round in feet was then
calculated from the following equation :

R = TiF — 150,

where 150 was the distance in feet from the T 4
tower to the gun position, and lay along the
line of reference.

The deflection D of the round in feet was the
perpendicular distance from the point of func-
tion to the line of aim. This distance in feet
was given by the approximate formula :

D = TiF (a — (3) .

Errors to Be Expected. There were several
factors affecting the accuracy of measurement

of the point of function. The influence of these
on the range of the shell is discussed below.

1. Errors in measurements of the locations
of the various towers. These points were sur-
veyed by a professional surveyor, and the meas-
urements were checked carefully. It was felt
that errors from this source were negligible.

2. Errors in measurement of angles at the
various towers. These angles could be measured
with an error of ±1 mil. Figure 48 shows

Figure 48. Graph showing errors in com-
puted distance from T\ tower to point of func-
tion for errors of 1 mil in angles measured at
various observation towers.

graphs of the errors in FiF2, T4F3, and 7\F4
to be expected when errors of 1 mil were made
in measuring the various angles. Inspection of
these graphs shows that by using the data from
the proper towers, the maximum error in the
distance from Ti tower to the point of function
for ranges from 1,000 to 12,000 ft is 13 ft.

3. There was also an error in subtracting 150
ft from the distance T±F in order to get the
range of the round, unless the point of function
fell along the line of reference. This error is
equal to 150 X sin a. Since a is a small angle,
the error from this source is insignificant.

4. The size of the nomograph constructed to
solve the equations was 24x36 in. Calculations
using this nomograph should be more accurate
than calculations using the conventional 10-in.
slide rule. Calculations showed that the nomo-

SECRE'

350

FIELD TESTING OF PROXIMITY FUZES

graph could be read with an error of less than
5 ft.

5. The method of calculating the deflection
was approximate, since the assumption was that
the sine of the angle (a — (3) equaled the angle.
This was nearly true, since the angle was small.
Errors in measuring the angles a and (3 would
also result in an error in calculating the deflec-
tion. Using an estimated measurement accuracy
of 1 mil, the maximum error in the angle
(a — (3) would be 2 mils. This would result in a

Figure 49. Houston film developer, Model 11,
Type K-3 in use.

maximum error of approximately 2 ft per
thousand feet of range.

Method of Obtaining Heights of Function

The height of function of the VT fuze was
determined by measuring the enlarged image
of a 16-mm motion picture film. Pictures were
taken from three positions: one behind the
firing point and two along the line of flight of
the shell. Bell and Howell Filmo Cameras,
Model 70-DA, were used, operating at maximum
speed of 64 frames per second. The camera at
the Ti tower pointed along the flight line and
was equipped with a 4-in. telephoto lens. At
the other two stations a 2-in. lens was used
unless the camera was pointed at a small enough
angle to the flight line to insure the function
being in a small field of view, in which case a
4-in. lens was used. The cameras were also
equipped with 1-in. lenses for photographing

(after each function) a blackboard giving the
round and fuze number. The operators started
the cameras for photographing functions when
the signal “camera” was given by the operator
in charge. This signal was given approximately
2 sec before the estimated functioning time.

The film was processed with a Houston Film
Developer, Model 11, Type K-3. Best results
were obtained by following the procedure in
reference 48. Ansco Hypan film was used. Fig-
ure 49 shows the film being processed in the
Houston Developer.

The Recordak Film Reader, Model C, was
used to obtain an enlarged image of the finished
film. The image was measured to the nearest
0.02 in. Magnification of the image was care-
fully checked by measuring the height of known
objects at known distances. Since the focal
length of the lens and the distance from camera
to function were known, the height of function
was computed, using the formula,

Height of function =

Measured height of image X distance to function
Focal length X K

where K is the magnification of the image. This
formula is accurate to 1 ft at a range of 4,400
ft. There was always at least one camera sta-
tion within this range. Hence the accuracy was
consistently within a foot of the height of
function if the picture showed the actual height
of function.

Fuze Flight Performance

A block diagram of the radio equipment is
shown in Figure 50.

Figure 50. Block diagram of radio apparatus.

A Zenith broad band receiver with a flat
frequency response of ±3 me was used. Two
dipole antennas were connected by 50-ohm
transmission lines to the receiver. One antenna
was used for the Globe-Union units, the other
for Zenith units. A Hewlett-Packard Model

THE FIELD TESTING OF MORTAR SHELL FUZES

351

200C audio oscillator supplied the 1,000-c note
used to obtain zero time. This 1,000-c note was
shut off at the gun position by a switch actuated
by muzzle blast from the gun. A mixer amplifier
fed both the 1,000-c note and the signal from
the receiver into the network at the same time.
This unit had very little gain and was used
primarily for mixing. A limiter, designed to
keep the signal at a constant level, was coupled
as shown to the Presto 8-K recorder. A micro-
phone connected into the recorder circuit made
possible the recording of voice announcements
associated with the firing of each round. All
recordings were made with a sapphire cutting
needle on Audiodisks. Before firing a given
round, the receiver was tuned to the expected
frequency of the unit listed on the test request
data sheet.

Coordinating the radio with the firing oper-
ation was carried out as follows. As the gun
operator began counting, the man at the tele-
phone in the radio laboratory counted aloud in
unison with him. This gave the other operator
time to start recording the 1,000-c note. As the
shot was fired, the note was cut off sharply by
the micro switch and the signal from the unit
cut in. The termination of the 1,000-c note gave
the zero time for the signal being recorded. A
check was kept on the relative strength of the
signal by observing the monitoring meter on
the recorder, designating the optimum signal
strength by the number 5. The volume of the
1,000-c note was always set to 4. The quality
of the signal was determined by ear, again des-
ignating by 5 the optimum quality. Both read-
ings were recorded on a data sheet for each
round. The radio operator timed the duration
of the signal with a stopwatch and at the end
of each signal recorded the round number on
the disk.

Determination of Generator Speed. To record
generator frequency, the signal received from
the fuze was fed to the horizontal plates of an
oscilloscope. The vertical plates were tied to-
gether and grounded. Best results were ob-
tained by picking up the signal from the moni-
tor of the recorder. The oscilloscope was then
photographed with a Bell and Howell, Model
70-DA, 16-mm motion picture camera from
which the shutter mechanism had been re-

moved and the spring drive replaced by a syn-
chronous motor giving a film speed of 15 in.
per sec. The camera and oscilloscope were en-
closed in a dark chamber.

Approximately a second before the shell was
fired, the camera was started. The 1,000-c note
from the audio oscillator was recorded until the
firing of the mortar operated the zero time de-
vice. The generator frequency was recorded on
the film from this time until the functioning of
the fuze.

Ansco Triple S Pan and Eastman Super XX
films were found to be the most easily read.
After processing, the films were marked at 15
in. intervals in order that readings could be
taken every second. Readings were also taken
at three points in the first second, since the
speed changed rapidly immediately after the
shell was fired. Reading of the films was done
on the Recordak Film Reader, Model C. Since
the generator had six poles, the generator speed
equals % of the frequency.

Although the equipment operated well, the
system was not completely satisfactory. The
1,000-c note was used as a standard to check
the apparatus and results were consistently
very good. However, the carrier from the fuze
was modulated by the plate voltage ripple and
also by the filament voltage. The filament modu-
lation gave the fundamental frequency and the
plate voltage ripple gave the first harmonic.
During the flight of the shell, the predominate
modulation was first the filament voltage, then
the plate voltage ripple, and finally the filament
voltage again at the end of the trajectory.
Hence, extreme care had to be taken to avoid
confusing the fundamentals with the harmonic.
Reporter units should have been used when ac-
curate generator speeds were needed.

Carrier modulation was also recorded on
phonograph records which were analyzed by
a method described in the Bibliography.13

Weather Data

The purpose of collecting weather data was
to obtain information pertinent to ballistic cal-
culations and also to determine any adverse
effects of weather on the VT fuze.

The direction and velocity of the surface
wind were determined, respectively, by a wind

SECRET

352

FIELD TESTING OF PROXIMITY FUZES

vane and an anemometer. To determine wind at
higher altitudes, a pilot balloon was released
each hour during firing and theodolite readings
taken on the position of the balloon. Reference
47 was used to convert these readings to an
average wind for each 1,000-ft zone. A ballistic
wind was then calculated by averaging these
zone winds weighted as to the time the projec-
tile was in each zone. Although no attempt had
been made to correct the ballistic data for wind,
the recorded ballistic wind could have been used
for this purpose.

The sky condition was recorded as overcast,
broken, scattered, or clear, depending on
whether the coverage of the sky was over 0.9,
between 0.5 and 0.9, between 0.1 and 0.5, or less
than 0.1, respectively. Standard weather bureau
terminology was used to describe cloud types.

Continuous daily recording of temperature,
atmospheric pressure, and humidity were made
by thermograph, barograph, and hydrograph,
respectively. Other pertinent weather informa-
tion, such as fog and precipitation, were also
recorded.

Field Test Reports

Immediately after a firing program was com-
pleted or at intervals during long programs,
data from the field were brought into the office
in Clinton and given to the computing group.
These raw data were then transcribed into a
calculation sheet.

Rounds fired were numbered consecutively,
and these round numbers were used to coordi-
nate all data. Data sheets were furnished to
the loading department, the gun station, radio
operators, and tower observers for recording
such information as the type of vehicle and
fuze, weight of complete round, charge, angle
of elevation, type of function, time of func-
tion, azimuth of function, strength and quality
of radio signal, and muzzle velocity. Specially
prepared forms were used to record the data
at the field.

A calculation sheet was made up to be used
with the nomograph described in the discussion
of ballistic data given in this section. On this
sheet, the columns were arranged to allow
orderly recording of the data for rapid com-
putation.

When the calculations were completed, they
were recorded on a field test report sheet. This
sheet contained the results of the test in a
compact form and included the principal data
submitted in the Field Test Reports. As used in
this report, the mean dispersion of a group of
variables was defined as the average of absolute
values of the differences between the mean
value of the variables and the particular vari-
ables.

The request for field tests included the fol-
lowing information: (1) originator of the test;
(2) the contact person representing the origi-
nator; (3) purpose of test ; (4) description and
conditions of test; (5) description of material
to be tested; (6) data required from the test;
(7) source of material ; (8) urgency of test ; (9)
relation of other test requests; (10) statement
as to whether originator's representative would
witness the test; (11) remarks; (12) approval
of test request by originator; (13) approval or
modification of test request by the director of
the field station.

Immediately upon receipt of a test request,
one person was assigned to study the test re-
quest and begin writing as much of the report
as was possible before the data were obtained
from the field station. This was necessary in
order to facilitate sending reports out on the
day the firing was done.

8.4.4 TJie Mathematical Calculation of
Mortar Shell Trajectories

Three methods for the computation of mor-
tar shell trajectories45 were considered, namely,
the Piton-Bressant procedure, a method espe-
cially adapted to the use of the quadratic air-
resistance law, and the method of numerical
integration as developed in 1918 and the fol-
lowing few years. It was found that existing
ballistic tables were inapplicable to the par-
ticular type of mortar shell in question, since
the tables give trajectories only for shells
whose ballistic coefficients are greater than or
equal to unity while these mortar shells have
ballistic coefficients of one-half or less.

The Piton-Bressant method requires a
knowledge of the initial conditions, that is,

SECRET

THE FIELD TESTING OF MORTAR SHELL FUZES

353

muzzle velocity and elevation of the mortar,
and of the range on the horizontal. This method
is the easiest of the three to apply, but it is sub-
ject to an inherent error of a magnitude indi-
cated by Figure 51. No particular assumption
is made about air resistance, the effect of which
is taken into account by the use made of the
measured range. The method permits the calcu-
lation of as many points on the trajectory as
may be desired, together with the time at each.

In the second method, it is assumed that the

Figure 51. Comparison of trajectory calculated
by Piton-Bressant procedure with exact tra-
jectory.

drag of the air is given by the expression cv2,
where v is the velocity of the shell at any time
and c is a constant whose value must be found
from field or wind tunnel measurements. For-
mulas were developed giving the coordinates
and time as functions of the inclination. By
means of these formulas, as many points on the
trajectory can be calculated as may be desired,
together with the time at each. This method is
capable of as much accuracy as the measured
data warrant. It was this method which was
used in calculating the “exact trajectory” of
the figure.

Since the type of mortar shell under consid-
eration has a ballistic coefficient less than unity
(corresponding to a relatively large air re-
sistance), the effect caused by the decrease in
air density with increasing height may be
appreciable on the trajectory as a whole. In
order to test this point the trajectory for a
muzzle velocity of 635 fps, an elevation of the

mortar of 65 degrees, and a ballistic coefficient
of 0.4284b was calculated by the method of
numerical integration using the tabulated
Gavre function first with account taken of the
change in air density with changing altitude
and then ignoring this change. The respective
ranges were found to be 5,033 and 4,901 ft.
The difference is 132 ft, or about 2.6 per cent
of the range.

Formulas were developed for the effects pro-
duced by small changes in initial conditions.
The method of differentials was used in this
connection. Application of these formulas was
made to the problem of the effect of wind on a
trajectory.46

8 4 5 Photographing Height of Function
of VT Fuze

Introduction

Experiments were performed for the pur-
pose of gaining more definite information on
the measurements of heights of function of the
VT fuze. The heights of function were calcu-
lated from photographic data taken at three
observation towers. To obtain the actual height
at which the VT fuze functioned, it was desir-
able to photograph the detonation of the tetryl
in the fuze ; this was the first indication of the
functioning of the fuze. During a certain pe-
riod, when potassium permanganate and mag-
nesium packs were not available to be placed
behind the tetryl pellet to make a puff after the
fuze functioned, black powder was employed
for this purpose. The functioning of approxi-
mately two hundred of the fuzes was photo-
graphed at this time by three cameras operat-
ing at 64 frames per second, and in no instance
did a flash appear on any of the films. Three
fuzes were then placed in the mortar shells in
the usual manner, except that the black powder
was missing, and were statically detonated.
Again, a flash was not recorded on any one of
the films. Hence, it was assumed that when
potassium permanganate and magnesium puffs
were used, the flash appearing on the film was
actually from the potassium permanganate

b Approximately the value for the M-43C shell with
the T-132 fuze.

SECRET

354

FIELD TESTING OF PROXIMITY FUZES

and magnesium and not from the tetryl pellet.
Potassium permanganate and magnesium was
used thereafter for making the puff, and a flash
appeared on the film for each proper function.

The disagreement between the heights of
function obtained from the photographs from
the three towers was greater than 1 ft in only 3
per cent of the cases and was never greater
than 2 ft. Thus, the method of measuring the
heights of the puffs seemed to be quite accu-
rate.

Experiments to Determine Whether
Actual Heights of Function Were
Being Photographed

Experiment I. The first evidence of function
should be the detonation of the tetryl booster
pellet. To discover if this detonation was ca-
pable of being photographed, 10 tetryl booster
pellets were statically detonated by taping each
pellet to an electric detonator and fired by
means of a hand magneto. Motion pictures
taken at 64 frames per sec showed a definite
flash. The explosions were also photographed
by a shutterless motor-driven camera. Meas-
urement of the length of a bright streak on the
finished film from this camera showed the dura-
tion of flash to be 0.02 sec.

Experiment II. Three M-43 inert loaded
shells were loaded with detonator and tetryl
booster pellet only. The tetryl was held in the
aluminum tetryl cup in a standard M-53 PD
fuze. These units were detonated statically as
in experiment I and photographed with three
cameras operating simultaneously. Photo-
graphic data showed a brilliant flash with an
average duration of 0.18 sec. The flash was
followed by a thin gray smoke.

Experiment III. Three M-43 inert loaded
shells were packed with detonator and tetryl
pellet. The tetryl was held in a brass tetryl cup
in a reject VT fuze. The units were statically
detonated and photographed with three cam-
eras. There was no evidence of a flash on the
finished film. A dense black smoke was the first
visible record of function. This possibly indi-
cated that the flash seen in using tetryl packed
in aluminum cups may have come from the
burning aluminum. The flash seen in the case
of tetryl alone was not present when the tetryl

was packed in brass, possibly because the
energy from the detonation was used in break-
ing and heating the shell.

Experiment IV. Eighteen units were loaded
as in experiment III and fired from the stand-
ard 81-mm M-l mortar. There was no evidence
of a flash when the functioning of these units
was photographed by three cameras. However,
smoke was plainly visible on the films.

Experiment V. Units loaded with a black
powder cartridge behind the tetryl cup showed
smoke as the first visible evidence of function.
In photographs of 200 units packed with the
black powder cartridge, there was no record of
a flash.

Experiment VI. Shells were ordinarily
packed with a cartridge of potassium perman-
ganate and magnesium behind the booster pel-
let. The functioning of this round photographed
nicely as a black point topped or surrounded
by a flash. The flash was found to photograph
well against a water background in all types of
weather. Considering the above experiments,
the conclusion is drawn that this flash was
caused by the potassium permanganate and
magnesium and not by the tetryl pellet. A com-
plete account of the preceding experiments to-
gether with photographs may be found in ref-
erence 43. Comparisons of black powder and
permanganate-magnesium spotting charges are
also reported in reference 23.

Comparison of Heights of Function
Obtained from Three Observation
Towers

The shutter on the cameras used in photo-
graphing the functioning of the VT fuze had
an opening of 204 degrees. When operating at
64 frames per sec, the shutter was open 0.0088
sec and then closed 0.0068 sec. Hence, the shut-
ter was open approximately % of the time and
closed % of the time. With three cameras oper-
ating essentially independently of one another,
the probability of missing the first evidence of
function was (%)3, or approximately Yli- If
the shell had a maximum approach velocity of
500 fps, it might have traveled as much as 3.4
ft while the shutter on one of the cameras was
closed. Since agreement among three camera
stations was consistently within a foot, it seems

SECRET

THE FIELD TESTING OF MORTAR SHELL FUZES

355

unlikely that the explosion was carried down-
ward with the same velocity as the shell. Any
error larger than a foot between readings was

point. Since the cameras were operating inde-
pendently, the detonation might have taken
place while one or more of the camera shut-

braze wires
TO CONTAINER

Cable guide

Steel tube

Steel cable

Base plate

Shear pin

Threaded base

Cover

Set screw

Stud for shear pin

10.

Time fuze

11. Parachute space

Figure 52. Type A-l mortar shell retrieving device.

invariably the result of faulty measuring or
computing.

Measurements were taken from the highest
point from which the explosion appeared to
emanate to the water below the functioning

ters were closed. However, the close agreement
of data from the three cameras indicated
that the point of function remained fixed and
visible for a sufficient length of time to be re-
corded photographically in all three cases.

SECRET

356

FIELD TESTING OF PROXIMITY FUZES

Furthermore, this close agreement seemed to
substantiate the fact that the results were accu-
rate. The only error present may have been the
result of the time lapse between the functioning
of the fuze and the appearance of visible evi-
dence of it. Considering the fact that tetryl has

Nor e- Use One Oram Base Char6_e

1. Adapter for M-65 fuze

2. Steel plate

3. Internal steel tube (half cylinders)

4. Shear pin

5. Threaded base

6. Time fuze

Figure 53. Type A-2 mortar shell retrieving
device.

a rate of detonation of thousands of meters per
second, the error introduced was less than 1 ft.

846 Parachute Recovery Devices
Introduction

During the development of the VT mortar
fuzes, it became apparent that some means was
needed to allow the fuzes to be subjected to

Figure 54. Type B fuze retrieving unit.

accelerations comparable with the accelerations
encountered in actual firing and still permit
further testing and inspection of these fuzes.
This was necessary so that damage to the fuze
caused by acceleration might be studied and
faulty or failing parts redesigned. The centri-
fuge furnished a partial solution to the prob-
lem. However, the accelerations available with

the centrifuge did not accurately simulate the
instantaneous acceleration encountered in ac-
tual firing. It was agreed that some means for
recovery of fuzes after they had actually been
fired and gone through part of a normal flight
was needed. The development of a recovery de-
vice was assigned to the University of Iowa.

Devices for parachute recovery not only of
the VT fuzes but of complete 81-mm mortar
shells were developed and put into production
and use. On the whole, these devices functioned
satisfactorily. The various types of devices,
their applications and use will be discussed be-
low.44

Type A-l Device

The first device developed consisted of a
tubular steel parachute container with
threaded base. The threaded base was screwed
into the mortar shell in the position normally
occupied by the shell fuze. This device was in-
tended for recovery of the complete projectile.
A fuze or other material of interest could be
mounted inside the shell body. The original de-
sign (type A) proved unsatisfactory and was
never used.

Figure 52 is a drawing of the type A-l de-
vice as used. Space was provided for a fixed
time (15 sec) powder train fuze (M-65) to
eject the parachute. The M-65 fuze was inserted
in the bottom of the container and the para-
chute was packed snugly against the fuze. The
cover was placed over the parachute and held
in position by three 0.081-in. half-hard brass
shear pins. The cover was fastened securely to
the container through two %-in. flexible steel
cables. One end of each cable was brazed to the
outside of the container. The other end of each
cable was passed down through the top of the
cover and back up through the cover from the
bottom. The ends were then brazed to the top of
the cover. A loop in the steel cables for attach-
ment of the parachute shroud lines was left on
the bottom side of the cover.

In operation, the M-65 fuze was initiated
when the shell was fired. After 15 sec of flight,
the powder train ignited a reduced (approxi-
mately 1 g) charge of black powder in the base
of the fuze and forced the fuze forward; this
sheared the pins retaining the cover and forced

SECRET

THE FIELD TESTING OF MORTAR SHELL FUZES

357

the parachute out. The whole shell was then
recovered on the parachute.

The principal difficulty encountered in the
use of this device was separation of the steel
cables from the container at the point where
they were brazed. This difficulty was prac-
tically eliminated by greater care not to over-
heat the cables in brazing. At the Clinton Field
Station, 80 of these devices were used to re-

mounting the M-65 fuze to operate the device
on the extreme front end of the assembly. The
base charge of the M-65 fuze was used to shear
half-hard brass pins and release the parachute.
Figure 53 is a drawing of the type A-2 device.

Type B Device

The type B device was developed very soon
after the type A device and was actually in pro-

Mortar shell

Part from M-lll or M-lll A-2

time fuze

Steel housing

Part from M-lll or M-lll A-2

time fuze

Part from M-lll or M-lll A-2 time fuze

Part from M-lll or M-lll A-2

time fuze

Steel tube (half cylinders)

Loop anchor for shroud lines

Safety wire

10.

Shear pin

11. Primer

Figure 55. Type C fuze retrieving device.

cover complete rounds of M-56 mortar shells.
Of the 80 used, 61 functioned satisfactorily.

Type A-2 Device

The type A-2 device was never built, though
drawings were prepared. It provided for secure
fastening of the parachute shroud lines to the
threaded base and eliminated the steel cables
used on the type A-l device. Approximately the
same size and shape container was used as for
the type A-l device. Provision was made for

duction first. It was built into a modified M-56
mortar shell. This device provided for mount-
ing the VT fuze in its normal position on the
shell.

The M-56 mortar shell was modified as fol-
lows. The nose of the shell was cut off just for-
ward of the three small bosses which touch the
barrel (actually right at the line between the
shell nose and the straight part of body). The
inside of the shell body was reamed to a stand-
ard size. A circular steel plate was pressed

358

FIELD TESTING OF PROXIMITY FUZES

down inside the shell to a depth of approxi-
mately 6 in. and secured by steel pins pressed
through the shell wall and into the plate. The
space forward of this steel plate was used for
the parachute and an aluminum piston in which
an M-65 powder train fuze was mounted. The
nose was fastened back on the shell body with
a short piece of steel tube pressed inside the
nose and held to the body with 3 half-hard
brass pins (0.081 in. in diameter). A loop of
steel rod brazed to the nose served as an anchor
point for the parachute shroud lines. The com-
pleted device presented the same external ap-
pearance as a standard M-56 shell. Figure 54 is
a drawing of the type B device.

When the shell was fired, the M-65 powder
train fuze was initiated and the time ring
began burning. After 15 sec of flight, the time
ring ignited the base charge (reduced to 1 g)
of black powder in the fuze. This forced the
nose off and expelled the parachute. The nose
with the VT fuze attached was brought down
on the parachute.

At the Clinton Field Station 45 of these de-
vices were used. Of the 45 tried, 44 functioned
satisfactorily. Several hundred of these devices
were shipped to Blossom Point, where they also
functioned satisfactorily.

All of the parachutes for types A, A-l, A-2,
and B devices were 36-in. Fortesan rayon cano-
pies with 100-lb rayon shroud lines. Some of the
canopies were white and some were dyed
orange to make them easier to follow.

Type C Device

One objection to the type B device was that
it was built into an M-56 mortar shell; conse-
quently, it was impossible to subject the VT
fuze to the accelerations desired. The design
was such that it could not be adapted to the
lighter M-43 shell with which the higher accel-
erations might be realized. Therefore, develop-
ment was begun on a completely new device
which would work equally well in the M-56 or
M-43 mortar shell. This device, designated as
type C, was still in the development stage when
operations were terminated. The type C device
was a complete unit which might be inserted in
either an M-56 or M-43 A-l empty shell after
removal of the adapter ring in the shell nose.

It provided for mounting the VT fuze in its
normal position. After the type C device was
inserted in either the M-56 or M-43 A-l mortar
shell, the shell presented approximately the
same external appearance as before. The device
was built into a steel tube approximately 6 in.
long and l3/4-in. inside diameter. Space was
provided for an adjustable time mechanical
time fuze, a parachute, and a nose ring with
threads for the VT fuze. The time fuze used
was actually part of two standard fuzes, the
M-lll or M-lll A-2, and the M-136. The timing
elements or clocks in these two fuzes were
identical in external appearance and differed
only in the rate at which the timing disk
turned. The body of the M-lll fuze and the
clock from the M-136 fuze were used. The hy-
brid fuze thus made up was modified so that
it was acceleration initiated instead of arming
wire initiated as originally. The delayed arm-
ing mechanism was removed completely. The
parachute used was a 30-in. Bemberg rayon
canopy with 40-lb rayon shroud lines. Figure
55 is a drawing of the type C device.

In operation the time fuze was initiated
when the shell was fired, and the base charge
was ignited at any desired time thereafter
(time was adjustable from 5 to 30 sec). This
forced the time fuze forward, pushed the nose
ring off and expelled the parachute. The nose
ring and the VT fuze were recovered on the
parachute.

At Clinton Field Station, 15 units were
tested. Of the 15 tested, five functioned satis-
factorily.

Recovery Procedure

In practice, rounds fired for recovery were
fired at an elevation of 75 to 80 degrees so that
the shell would be traveling slowly when the
parachute opened. Actual recovery of shells or
fuzes was somewhat complicated by the fact
that firing was done over water and recovery
was by boat.

Ordinarily two boats were sent out. They
stood by out of the line of fire until after the
parachute opened. Usually the men in the boats
could see the parachute open and get into posi-
tion to pick it up very soon after it hit the
water. When the men in the boat did not see the

SECRET

THE FIELD TESTING OF MORTAR SHELL FUZES

359

parachute, flag signals from shore were used to
direct the boats toward the point where the
parachute was expected to fall.

Usually the parachute hit in such a manner
as to trap air between the water and the

Figure 56. Breech-loading mortar, 81 mm.

canopy and remained afloat for several min-
utes. If the parachute sank, aiming circle bear-
ings from two towers ( Tx and To ) were taken
on the point where parachutes sank. Use was
made of these data to recover several shells
which otherwise would have been lost. On some

occasions the wind was such that the para-
chutes were carried over land. This made re-
covery much less difficult. In fact, some were
blown back to the firing point and were caught
without striking the ground.

Recovered fuzes were dried as thoroughly as
possible with warm air blasts before being re-
turned to the interested parties for examina-
tion.

847 Recovery by Use of Breech-Loading
Mortar

The firing of shells into a rectangular trough
filled with cotton waste was another very satis-
factory method of recovery. The horizontal
breech-loading mortar, shown in Figure 56,
was used. The trough was 20 ft long. The con-
struction of a metal detector for use in locat-
ing the shell within the trough was considered
but was found to be unnecessary. The heat de-
veloped within the waste proved to be a suffi-
cient indicator of the trajectory of the shell
within the waste for rapid recovery by hand.

SECRET

Chapter 9

ANALYSIS OF PERFORMANCE

91 INTRODUCTION

911 Purpose

IT is the purpose of this chapter to present
an analysis of the performance of variable-
time [VT] fuzes based on results obtained
mainly by those methods of field testing de-
scribed in the preceding chapter. Where pos-
sible, these results are compared with predic-
tions of performance based on the theory of
operation of the fuzes and on the characteris-
tics of the fuzes obtained in the laboratory
described in earlier chapters.

9,1,2 Sources of Data

Classification of Tests

Data on the field performance of the fuzes
are not limited to the proving grounds or meth-
ods described in Chapter 8. As pointed out in
Chapter 5, valuable data were obtained through
the courtesy of various military agencies. Field
tests may be classified roughly as follows:

1. Experimental tests performed during the
course of development of a fuze. For any fuze
that reached the mass production stage, the
results of development tests are of no more
than historical interest and are not given here.

2. Acceptance tests. For many fuzes, the
acceptance tests performed by Army Ordnance
provide voluminous data obtained under stand-
ardized test conditions. The conditions of the
acceptance tests are described in an appendix
to this chapter, and considerable use is made of

a This chapter was prepared by T. N. White, Jr., with
the assistance of Rachel Vorkink, Alan Leiner, and
Gladys Rabinow, Ordnance Development Division, Na-
tional Bureau of Standards, and Paul F. Bartunek,
Rosemarie Kilker, and David Fisher. In addition, H. F.
Stimson, of the Heat and Power Division, National
Bureau of Standards, prepared the sections dealing with
afterburning, and Walter G. Finch, former captain
in the VT Fuze Detachment of the Ordnance Depart-
ment, prepared Section 9.6 on operational use. Captain
Finch is now a graduate student at Johns Hopkins Uni-
versity. The summary, Section 9.7, was prepared by the
editor.

the results under the title, “Performance under
Acceptance Test Conditions,” in various sec-
tions of the chapter.

3. Experimental tests performed with pro-
duction fuzes or with fuzes closely approxi-
mating production design. These tests are of
particular interest in that they include experi-
ments to determine fuze performance under
various conditions that are of importance in
Service use but which are different from the
acceptance test conditions. In addition to tests
performed at the proving grounds, described in
Chapter 8, this category includes certain im-
portant Service tests performed by military
proving grounds.

The results of the above types of tests are
given in the various sections of this chapter
where the performance of the pertinent fuze is
under discussion. Reports on the results of com-
bat operations with VT fuzes are summarized
in a separate section. As would be expected,
these important results are of a qualitative
nature, mostly statements of the judgment of
observers working under very difficult condi-
tions. Such results are not susceptible to quan-
titative analysis, and no attempt was made to
subject them to such treatment.

Test Reference System

The inclusion of round-by-round results even
in the microfilm supplement of this report is
impractical. Reports on approximately 2,000
tests were prepared by the Ordnance Develop-
ment Division of the National Bureau of Stand-
ards alone. There is, therefore, included in the
microfilm supplement a set of tabular sum-
maries of the results of individual tests identi-
fied by test number, and reference is made to
these test numbers to show the sources of the
data presented in the text. These summaries
give the most important conditions of each test,
and also references to the detailed report on
each test. (Such summaries are not available
for Army rocket fuze tests ; in this case, refer-
ence is made directly to the detailed report in
order to give positive identification of the

360

| SECRET

INTRODUCTION

361

source material.) Some of these summaries
cover tests performed by military agencies, and
in some cases the data were provided through
courtesy of the military agency in advance of
the official report of the agency. In all cases
where an official report was available, refer-
ence is given to the official report. Every effort
has been made to attain accuracy in these sum-
maries, but it should be understood that there
is no implication that the military agencies con-
cerned are in any way bound by the results
given or by the interpretation of the results
presented in this report.

9,1,3 Description of Performance

The ideal representation of the perform-
ance of a fuze would be a diagram or model
showing the frequency of bursts in space under
each testing condition. A schematic one-dimen-
sional representation is shown in Figure 1.
This diagram is typical of bomb fuze perform-

Figure 1. Schematic distribution of VT fuze
functions along trajectory.

ance, except that the region of proper func-
tions has been expanded greatly (relative to the
total length of trajectory) in order to show
better the form of the proper function distri-
bution.

The data from most tests were much too
limited to yield a representative frequency dis-
tribution. The bursts that occurred in each test
were therefore classified as proper, early, dud,
etc., according to the position or time of oc-
currence. This method gave the score of the
test. For certain classes of burst, particularly
the proper bursts, the average position and
some measure of scatter were usually esti-
mated.

Test programs such as acceptance tests,
which yielded large masses of data, showed
that there was no sharp dividing line between
the different classes of burst. It was not always
possible even to distinguish between a burst at
the end of the flight and a dud. The method of
classification of bursts was therefore to some
extent arbitrary. For practical purposes this
uncertainty is a matter for little concern. In-
spection of Figure 1 shows that the proper
burst score, which is the most important score,
is little affected by the position of the limits
within which bursts are classed as proper, pro-
vided these limits are placed well out on the
“tails” of the proper burst “hump.” Generally
speaking, there was rarely any serious diffi-
culty in distinguishing between proper func-
tions and malfunctions, except in the very
earliest stages of testing of a basically new
fuze.

In the foregoing discussion it is tacitly as-
sumed that a function should be classed as
proper if the function is due primarily to in-
teraction between the fuze and the target. In
general, this is the criterion that was used in
experimental testing. There is, however, an-
other criterion that was used to some extent in
acceptance testing. This criterion is derived
from an assumption as to the space limits with-
in which an air burst may be regarded as useful
in inflicting damage in Service applications. In
a few cases the limits so set were a little severe
and were later broadened. The effect of such
changes on aggregate acceptance test scores
was, however, practically nil, and no attempt
has been made to revise old scores in order to
reduce all to a rigorously equal basis. It is also
true that the differences between the two cri-
teria mentioned above had a negligible effect on
the estimates of performance.

Throughout this chapter, the terms burst and
function are used interchangeably, as is the
case in most of the reports of the Division 4
NDRC Central Laboratory, and it is only in
rare instances that there is any important dis-
tinction between the two. However, for the sake
of exactness it appears worth while to empha-
size that it is the position of a flash or smoke
from a spotting charge or high-explosive [HE]
load that is actually estimated. The position at

362

ANALYSIS OF PERFORMANCE

which the fuze functions cannot be measured
directly in dynamic tests.

914 Evaluation by Field and

Laboratory Testing

Before considering the results of field tests,
it is important to realize that the evaluation
of any particular fuze design was not based
solely on its field performance, although satis-
factory field performance was essential to the
final acceptance of a design. Dependence on
laboratory data was particularly important for
fuzes in the earlier stages of development. It
is safe to say that if it had been essential to
obtain statistically convincing proof of the
value of every design change by means of field
tests, very few fuzes would have reached the
production stage during World War II. The
success of each fuze development was depend-
ent on the soundness of the engineering theory
of action of the fuze (which involved various
simplifying assumptions), the validity of labo-
ratory testing conditions (which could only
approximate field conditions), and field testing
(which was limited by both the time and labor
required to build fuzes, and by difficulties in
duplicating service conditions).

Although the preceding remarks apply pri-
marily to developmental work, they have an
important bearing on the evaluation of produc-
tion fuzes. In examining the data on field per-
formance, the following characteristics will
frequently be noted. 1. The data are volumi-
nous for acceptance test conditions but, for
many fuzes, quite scanty for other conditions
(e.g., other projectiles or velocities). 2. There
is frequent evidence of statistically significant
but unexplained variations between results
supposedly obtained under the same conditions
or between observed and predicted perform-
ance.

Although a very large amount of field test-
ing was done, it was impossible, under the con-
ditions that existed, to test all fuzes under all
important conditions. The available facilities
had to be reserved for the most urgent prob-
lems. The value of engineering predictions
based largely on laboratory data, as attested by

the development work, and by the performance
of the fuzes that were tested under a variety of
conditions, provides a reasonable assurance
that certain gaps in the pattern of field data
need not cause great concern.

With regard to the statistically significant
variations, it will be noted that they are in most
cases too small to be of practical importance in
the military use of the fuze. In some cases,
explanations might be given in terms of the
approximations involved in the theory, labora-
tory, or field testing of the fuzes. These expla-
nations are usually mentioned only in those
instances where the discrepancies are consid-
ered to be of a practical magnitude.

915 Terminology

Fuze Nomenclature

In presenting the results of acceptance test-
ing, the designations of Chapter 5 are used.
The results of tests performed under other con-
ditions were obtained in many cases with both
production and pilot-production fuzes (see Sec-
tion 9.1.2, class 3) . Where a mixture occurs, the
simplest Army Ordnance designation is used
(e.g., T-51 for T-51, T-51-E1, T-51-E2, or
M-166). The exact composition of the group
is determinable through the reference system
for tests.

Manufacturers’ names are used to some ex-
tent on account of certain differences in per-
formance of the same fuze produced by differ-
ent manufacturers. Almost all of the important
differences in general quality of performance
were associated with fuze design rather than
with manufacturer. However, in the study of
the effect of certain factors on fuze perform-
ance, e.g., effect of altitude of bomb release
on burst height, it is sometimes necessary to
distinguish between manufacturers in order to
obtain a strictly valid test of the particular
factor under consideration.

For the sake of simplicity, certain obvious
abbreviations are used for manufacturers’
names.

Type of Function

The most important terms and abbreviations

sec:

FUZES FOR 4.5-IN. ARMY ROCKETS

363

are given in Chapter 5. Additional terms and
comments on usage in the older literature ap-
pear in the appropriate sections of this chapter.

Errors

Values given for the mean distance to a burst
are calculated where possible from photo-
graphic data obtained by methods described in
Chapter 8. Only where photographic data were
not available are visual estimates used. Only
in acceptance testing is a large quantity of vis-
ual data (camera obscura method) involved.
The discussion of systematic observational er-
rors is covered in Chapter 8. The values of
standard deviation of a distribution and stand-
ard error of the mean that are given in Chap-
ter 9 are calculated from individual observa-
tions of the test (or tests) involved. Except
where stated, no attempt is made to allow for
sources of systematic errors. In most cases
these measures of precision are utilized only in
the comparison of mean values obtained under
like observational conditions, so that the sys-
tematic component of error is balanced out.

Methods used in estimating the probability
of fortuitous differences are those available in
standard modern texts.98

92 FUZES FOR 4.5-IN. ARMY ROCKETS

921 Introduction

This chapter section deals with the perform-
ance of T-5 and T-6 fuzes for the Army 4.5-in.
rocket.

There was a large amount of testing that
provided information on the performance of
both the T-5 and the T-6 fuzes. For this reason,
the discussion of the performance of the two
fuzes is preceded by a section on tests that pro-
vided basic information on the performance of
both fuzes.

It should be noted that in target firing at
Corncake (Fort Fisher, N. C.) Proving Ground
(700 ft from launcher to target) 0.4-sec
SW-200 switches were used, while at Blossom
Point (1,200 ft from launcher to target) 0.7-
sec SW-200 switches were used.

The following terminology is used in this

chapter. In firing from the ground, or from
a plane, for function on approach to a ground
or water surface :

E = Early function, a function within 5 sec
of firing in the absence of any legitimate target.

M — Middle, or mid-flight function, a func-
tion that occurs more than 5 sec after firing but
too soon to be regarded as a proper function on
approach to the ground or water surface.

P = Proper function, a function that occurs
on approach to the ground or water surface
within limits of height that experience has
shown to be reasonable for normal operation of
the fuze. The term Pw or Pg may be used to
indicate that the function occurred over water
or over ground, respectively.

D = Dud.

N = Number of fuzes fired.

For the benefit of anyone who has occasion
to refer to source material, it should be noted
that at times the following terminology has
been used: W for Pw; A (approach) for P. In
firing at short range at a mock target (as in
acceptance testing), the following terms have
meanings different from those defined above.

E = Early function, any function occurring
before the target at a distance so great that a
proper function would be highly improbable.

P — Proper function, a function that occurs
at a position such that, judging from experi-
ence, it may reasonably be attributed to normal
interaction between the fuze and the target.

L = Late function, a function that occurs
after the region of P’s.

I = Impact function, one that occurs on strik-
ing the surface. In the source material and
reports on target tests, the term T (target)
has been used extensively for P. Also, L has
been used for spontaneous late functions, with
functions attributed to the passage over a land-
water boundary or to approach to water classi-
fied as B or W, respectively. In most cases the
number of functions in these classes was so
small that subdivision of the L class appears
unwarranted for the present purpose.

Although scores and scoring methods can be
discussed best in connection with experimental
results, a few preliminary remarks are desir-
able for purposes of orientation. Proper func-
tion scores are in all cases given as the number

5ECRET

364

ANALYSIS OF PERFORMANCE

of proper functions expressed as per cent of
the total number of fuzes fired, excluding from
consideration those rounds that did not have a
fair chance, e.g., rocket blowups or rounds that
passed outside of the region of action of a
target.

Early function scores may be reckoned in
different ways. For example, in the extensive
special studies on the causes of early function-
ing, duds provided no information, and it was
customary to exclude them from consideration
in calculating the early function percentage.
Fuzes that had functioned early could not again
function in mid-flight, so it was customary to
express the middle function score as a percent-
age of middles plus propers, usually excluding
duds as in calculating the early function score.

These scoring methods, which were suitable
for basic studies, are not directly interpretable
into performance of T-5 and T-6 fuzes. The
interpretation will be discussed after the data
have been presented. At this point it is merely
noted that a middle function would probably
appear as a proper function in the T-5 appli-
cation (provided the rocket passed reasonably
close to a target). Further, since no correla-
tion was found between middle and early func-
tioning and since early functioning does not
occur in the T-6 because of reliability of the
arming mechanism, the early functions may be
disregarded in estimating the performance of
T-6 from many tests in which the SW-200
switches were used.

92 2 Tests Yielding Basic Information
on Both Types of Fuzes

Early and Mid-Flight Functioning

On the basis of the principal causes of mal-
functioning of the T-5 and T-6, the random
functions have been divided into two classes,
early and middle, as defined in the preceding
section.3*5*17’32 Although the line of demarca-
tion (5 sec) is somewhat arbitrary, it will be
seen from the following discussions of the two
types of functions that from a practical stand-
point this division is quite satisfactory.

Early Functioning (Afterburning). On
many rockets a radio proximity fuze is handi-

capped by malfunctioning which is due to aft-
erburning of the rocket propellant (cf. Section
2.13). When flame issuing from the rocket
nozzle is ionized, it increases the effective
length of the rocket as an antenna, and makes
a change in the radiation impedance. Sudden
changes in the length of the flame make rapid
changes in the radiation impedance and hence
produce the same effect on the fuze as the
normal target. For this reason fuzes are gen-
erally constructed so that arming is not com-
pleted until after the primary burning flames
from the rocket have ceased. Frequently, how-
ever, there is a burning following the primary
burning, known as afterburning, and it is well
established that this afterburning is one of the
major causes of malfunctioning of rocket
fuzes. Static experiments were performed to
establish the correspondence of the fuze per-
formance with the properties of these after-
burning flames. These experiments also showed
that flame sometimes was present without
pulses but that triggering pulses were not pres-
ent without flame. Therefore, in order to avoid
an excessive proportion of malfunctions, it is
desirable to eliminate or control afterburning
from the motor.b

For the best performance of the rocket, the
pressure within the motor should decrease only
slightly during the primary burning. The pres-
sure within the motor, however, is strongly
dependent upon the surface area of the propel-
lant which is burning, so that after the surface
has decreased by a small amount, the pressure
has decreased by a larger amount. In order to
maintain the pressure, the shape of the pro-
pellant used in rocket motors is such that its
burning surface remains nearly constant
throughout the primary burning. In the 4.5-in.
Army rockets, which use solvent-extruded Bal-
listite, the grains of propellant are tubular and
the burning proceeds both from the outside of
these tubes and from the inside at the same
time. The surface on the outside of the grains
decreases and the surface on the inside of the
grains increases at essentially the same rate,

b Early attempts to eliminate malfunctioning during
the secondary burning period took several different
forms. Plugs to close the nozzle after the main blast, or
“sweeps” to remove residual powder, were tried. None
of these methods gave satisfactory results.

SECRET

FUZES FOR 4.5-IN. ARMY ROCKETS

365

so that the total surface remains nearly con-
stant, except for the shortening of the length
of the grains.

When burning has proceeded until the web
of Ballistite has been burned through over a
considerable portion of the grain, the surface,
and therefore the pressure, is reduced to such
an extent that primary burning is no longer
supported. In the Army 4.5-in. rocket the pri-
mary burning stops at about 0.2 sec. At this
instant, the temperature of the remaining Bal-
listite is very little greater than it was before
the primary burning started, because the sur-
face of the Ballistite, which was receiving heat,
was being consumed rapidly. After the primary
burning has stopped, and Ballistite is not being
consumed rapidly, the residue of Ballistite is
heated by radiation from hot metal parts with-
in the motor. A secondary low-pressure burn-
ing begins then and continues until all the re-
maining Ballistite is either consumed or
ejected.0 This seldom lasts more than 4 sec.

The products of combustion of the Ballistite
during the primary burning consist of some
inert gases such as C02 and N2 and also some
incompletely burned products such as CO and
H2. The incompletely burned gases, when mixed
with the oxygen of the air outside the rocket,
are probably the cause of the luminous flame
during the primary burning. During the sec-
ondary burning, there is probably an even
greater proportion of flammable gases which
can combine with the oxygen of the air to pro-
duce luminous flame. It seems to be a matter of
chance, however, whether these flammable
gases on issuing from the rocket nozzles will
ignite or not. The constituents of the Ballistite
also determine, to some extent, whether these
gases ignite or not ; for example, Ballistite
salted with 1.5 per cent K2S04 has much less
afterburning than the unsalted Ballistite. The
expansion ratio in the rocket nozzles may also
have a determining effect upon the temperature
and consequent ignition of these gases.

Since it was recognized that the burning of
the residual Ballistite was causing malfunc-
tioning of the fuzes, some method was sought

c Insulation of various sorts was applied to trap wires
and inside of motor but no improvement in performance
was noted.

to consume this Ballistite before the fuzes
armed. It was suggested that a mixture of
nitrate and picrate salts could be found which,
when mixed with a suitable binder and pressed
into pellets, would burn for 0.5 sec. Such pel-
lets, when added to the propellant charge, were
expected to consume the slivers of Ballistite
and entirely eliminate all flammable material
from the motor chamber before the arming
time of the fuze. Section H of Division 3 of
NDRC in Washington, and Division 8 of
NDRC at Bruceton, Pennsylvania, cooperated
in this search and developed “maintainer pel-
lets,” or “purge pellets” (as they were com-
monly called), for this purpose.

Tests had shown that when normal Ballistite
was used, nearly 70 per cent of the fuzes func-
tioned before 5 sec. Such functions were called
early functions. The addition of certain pellets
reduced the number of early functions to less
than 20 per cent. Contrary to expectations,
however, afterburning was not completely
eliminated. Firings at night showed that after-
burning often persisted continuously for an
even greater time when pellets were used than
when the standard charges were used, although
the afterburning was not so brilliant as it often
was without pellets. It is possible that the effec-
tiveness of the pellets was due in part to the
greater steadiness of the afterburning and in
part to its reduction.

At about this time, during the development
of these pellets, a variation in performance of
the rounds without the pellets was noticed
which was at first attributed to the particular
lot of propellant which happened to be loaded
in the motors. It was proven later, however,
that the lot of propellant had little, if anything,
to do with the fuze performance, but that the
variations in performance were almost en-
tirely dependent upon the interior metal parts
of the rocket motor. It was found, for example,
that the M-9A1 rocket motor gave about half
the percentage of early functions which the
earlier M-9 motor had given and the M-9A2
motor gave an intermediate performance.

The most striking discovery was that with
a double supporting ring at the base of the
trap on which the propellant was loaded there
were about 67 per cent early functions, where-

SECRET

366

ANALYSIS OF PERFORMANCE

as with a single supporting ring at the base of
the trap there were only about 25 per cent early
functions. Later a scalloped ring at the base
of the trap was developed by the Army as a
standard for this rocket, and with it there were
only about 18 per cent early functions. The
early functioning on the rockets was reduced
by this trap to about the same extent as by the
pellets on the double-ring traps.

Subsequent experiments were made, using
single-wire traps of different weights, and using
varying amounts of metal near the nozzle end
of the motor, but no explanation of the marked
effect of traps has yet been found. Furthermore,
extensive measurements of nozzle sizes, ratio
of length to throat diameter, were made when
the performance of later models of the M-9-
type, with variations in nozzle dimensions, were
found to give improved performance. These,
however, shed no light on the problem.

Simultaneously with the development of pel-
lets, work was done on salted powders. Some
of the more successful ones reduced the per-

Table 1. Early function! scores* for T-5 fuzes
on 4.5-in. Army rockets. Elevation is 60° or
greater unless otherwise noted.

Trap ring

Motor

%Et

Load: regular double-base propellant

Single wire

M-9

Single wire

M-9A1

Double wire§

M-9

443

175

Double wire

M-9A1

Scalloped

M-9

Scalloped

M-9A1

107

Load: regular plus 10 pellets

Double wire

M-9

104

Load: salted powder

Double wire

M-9

* Note: The better scores on the M-9A1 motors may be attribut-
able to the rotation of this projectile, which is brought about by
hand crimping of the fins.

t Including middle-function performance; see following section for
discussion of middle functioning.

t Disregarding duds, i.e., 100 E/(E + M + Pw) .

§ 58 of these were at 45° quadrant elevation, 27 E, 3 M, 28 W.

centage of early functions, with double-ring
traps, to about 20 per cent. When a mixed load,
part salted and part unsalted, was used, inter-
mediate scores were obtained.

It should be mentioned that investigations
were made of the effect of powder weight, of
motor velocity, and of the dampness and tem-

perature of propellant. In certain cases some
change in the time distribution of earlies was
noted, but there was no appreciable dependence
of functioning scores on any of these factors.

In Table 1 scores for various motor, trap,
and propellant combinations are given. Time
distribution of earlies is shown in Figure 2.

T-5 fuzes (high-angle firing) : A, with double-
wire ring at rear of trap; B, with single-wire
ring at rear of trap ; C, with purge pellets.

Middle Functioning. The principal known
cause of the random functions with T-5 and T-6
fuzes classified as middle (after 5 sec) is faulty
fin assemblies. Rounds fired, at 30-degree ele-
vation, on 4.5-in. rockets with nonlocking fins
give approximately 70 per cent such malfunc-
tioning. By proper modifications (crimping) to
insure locking of the fins in the open position,
this percentage may be reduced to about 20 or
less. Such expedients as brazing and welding
the fins in the open position also lowered the
middle-function score, in some cases quite mark-
edly. However, since rigid fins prohibit the use
of a smooth-bore tube as a launcher, emphasis
was not placed on their development.

Following the discovery of the strong depend-
ence of early functioning on the type of trap
used, a comprehensive study was made of ex-
isting results to determine any relation that
might exist between middle functioning and
known variables (other than fin assemblies)
such as motor traps and propellants. No de-
pendence of middles could be found upon (1)
trap construction, (2) type of propellant, in-
cluding pellets and salted powders, or (3) fre-
quency band of the fuze (i.e., Red, Yellow, or

SECR

FUZES FOR 4.5-IN. ARMY ROCKETS

367

Green). Here it should be mentioned that ex-
periments were performed to test the effect
upon fuze performance of loose joints, both
between shell and fuze and shell and motor. It
was shown that any reasonable looseness of
these joints would not produce an increase in
middle functioning.

When results were sorted according to manu-
facturer, however, statistically significant dif-
ferences were found as follows :

The possibility of such functioning was in-
vestigated in a test where 60 T-5’s, mounted on
HE-loaded 4.5-in. rockets, were fired in 12 sal-
vos of 5 each. The nominal time interval be-
tween successive rounds was 0.1 sec. The
launchers, 10 ft long, were mounted in parallel
with a space of 10 in. between centers and at
an elevation of 50 degrees. The self-destruction
[SD] switch of one fuze in each salvo, usually
that in the middle position, was set to go at

Mfr.

Overall
% middle

% middle in
30 sec*

No. of rounds
on which %
is based

14.3

12.1

29.1

24.6

26.0

22.0

938

598

* Thirty seconds is approximately the flight time for maximum
firing elevation (42 degrees) prescribed by the Army.

Plots of “per cent still good” versus time,
were made and found to take the form of ex-
ponential curves. When these curves were ex-
trapolated back into the early-function period,
it was found that from 5 to 10 per cent of the
malfunctions scored as earlies should probably
be attributed to the middle-function phenom-
enon.

A very satisfactory reduction of middle func-
tions was found in units which had survived
rather violent “shaker” testing in the labora-
tory (see Section 7.4). Forty-one units not sub-
jected to such testing gave 22 per cent middles,
while 27 shaker-tested units fired under similar
conditions gave no mid-functions.

A summary of representative middle-func-
tion performance is given in Table 2. Figure 3
shows the time distribution of middle functions
for some 64 rounds on Revere M-9 4.5-in. rock-
ets with nonlocking fins (30-degrees quadrant
elevation [QE] ) . This distribution, which
shows no particular bunching of functions at
any specific time interval, is typical of the T-6
middle-function performance.

Mutual Interference6

So far the discussion of random functions
has been confined to rounds fired singly. It is
evident that in multiple firing a serious prob-
lem might arise from sympathetic functioning,
i.e., the triggering of one fuze by the function-
ing of a neighboring fuze.

Table 2. Middle-function scores for T-5 and T-6
fuzes on 4.5-in. Army rockets.

Fin type

Per

cent

middle*

Quad-

rant

No. of eleva-
rounds tion
mid and (in

proper degrees)

Fuze

mfr.

T-6

Locking, factory
crimped

25-40

Nonlocking

106

A, B

Hand crimped
(locking)

Crimped and
brazed

A(T-5)

Welded, single
thickness

Welded, double
thickness

T-5, shaker tested
Hand crimped

Hand crimped

T-5, controls for shaker tested

Hand crimped 22 41

* %M = 100 M/(M + W).

2.5 sec, so that one fuze would be certain to
function before the normal time for SD func-
tioning (usually between 6 and 12 sec).d

Results of the test were somewhat compli-
cated by several factors.

1. Not more than half the fuzes set for an
early SD time functioned during the desired
period. (This meant that only a very small
amount of data covering the useful portion of
the T-5 flight could be obtained.)

d T-5 fuzes normally come equipped with this type of
switch although discussions in previous sections pertain-
ing to middle functioning of the T-5 fuzes were confined
to results with fuzes in which the SD switch had been
shorted out.

▼secret

368

ANALYSIS OF PERFORMANCE

2. Variations in initial velocities and irreg-
ularities in firing intervals made distances be-
tween rockets in flight quite uncertain.

3. Times to function as determined by stop-
watch could not be considered very accurate.

In order to make allowance for errors in
timing and to provide some means of analyzing

the arming switch. In most cases the rotation
of the projectiles was brought about by the
deformation of fins during the crimping proc-
ess. A series of field tests confirmed laboratory
results on the delay or prevention of arming7’ 16
due to rotation. A general conclusion was that
the effect became serious if the fins were tilted

Figure 3. Distribution of functions in mid-flight, T-6 on M-9 with nonlocking fins. Elevation: 30°.
Eash dash on trajectory shows approximate position of a function.

the data, the following method was used. To
each round there was assigned a 0.4-sec inter-
val spanning the given time to function. A rea-
sonable measure then of the presence of sympa-
thetic functioning was a comparison of the
number of overlapping intervals within salvos
and those between salvos.

Statistical analysis showed good agreement
between expected (fortuitous) and observed
members of overlapping pairs within and be-
tween salvos. This indicated that no appreci-
able sympathetic functioning occurred.

Miscellaneous

Spin Effect on Arming. The SW-230-type
arming switch, when used on nonrotating pro-
jectiles, is very reliable. When, however, this
switch is subjected to rotation in excess of a
certain minimum speed, faulty performance is
to be expected. Laboratory testing has shown
that for rotational speeds up to 600 rpm normal
functioning occurs; above 900 rpm, the switch
does not arm at all ; and between these two ex-
tremes there is an increasing time required to
complete arming.13 The effect was of practical
importance because a slight twist on the fins
of an M-8 rocket might produce enough rota-
tion to interfere with the proper operation of

more than 2 degrees from their proper position.

Rain Effect. The T-5 (or T-6) fuze when
fired in moderate or heavy rain cannot be de-
pended upon to ride through to proper func-
tion. The triggering pulses from impact with
the drops can be significantly reduced, however,
by the use of Lucite caps cemented over the
conical surface of the fuze. The following gives
a comparison of function scores for rounds
fired during rainfalls of comparable intensity0
with and without such “rain caps.”

E M Pw D
With Lucite caps 10 8 1

Without caps 5 13 2

Photographic measurements of function
heights indicated no appreciable effect on sen-
sitivity from the Lucite caps.

923 Performance of T-5 Fuzes

Safety and Arming

General Considerations. The arming switch
of the T-5 is so designed that arming occurs

e Data on frequency and size of drops were obtained
by exposing pieces of specially prepared cloth to the
rain for measured intervals of time. Where water hits
this cloth a permanent colored spot is produced.9

SECR

FUZES FOR 4.5-IN. ARMY ROCKETS

369

at a definite time after the end of burning. The
distance from the launching point to the point
of arming is therefore obtainable by adding to
the burning distance the product: (mean veloc-
ity during switch operation) X (time of switch
operation). As the temperature of the rocket
propellant is increased, the burning distance
decreases and the peak velocity increases. The
arming distance of the fuze is therefore a func-
tion of temperature.

When the rocket is launched from a plane,
the distance that is of interest is the distance
from plane to rocket at the time of arming. In
general this distance will be less than the dis-
tance determined in a ground launching test at
the same temperature. This decrease in distance
is due to the greater air drag on the rocket,
which travels at a higher speed when launched
from a plane. This statement is true in cases
of firing at moderate altitudes. In case of firing
at high altitudes, there may be a compensating
effect due to the higher efficiency of rockets in
rarefied atmosphere. The arming distance of
the T-5 is therefore a function of the tempera-
ture of the rocket propellant, the speed of the
launching plane, and its altitude.

Sufficient data are not available for exact
estimation of the effects of these factors on
arming distance. Approximate calculations in-
dicate that the effects can be neglected, for
practical purposes, under a fairly wide variety
of conditions. The possibility that they might
be of importance under extreme conditions
should not be disregarded.

Switch Reliability. 1. Failure to arm. Spe-
cific data on failure to arm, as such, are not
available. However, dud scores in acceptance
testing establish a reliable measure of the upper
limit of SW-200 switch failure. The overall dud
score for 4,334 rounds was 3.6 per cent. It is
reasonable to assume, therefore, that something
less than this percentage of switches failed to
arm.

2. Safety and lower limit for arming. Data
on time and distance to arming from direct
measurements on units set to function on arm-
ing with 0.7-sec switches, are very scanty.
Again, reference may be made to acceptance
results to establish lower limits. In Figure 4
is given the distribution of 226 early functions

(Blossom Point data, all on M-9) in terms of
distance from the launcher. From this curve
it may be seen that less than 1 per cent of the
fuzes had functioned at the 550-ft point and
none at 525 ft. Although there is no certainty
that some fuzes had not armed before the 525-ft
mark from the standpoint of safety, it is rele-
vant to emphasize the fact that no functions
were observed before this point. For standard

Figure 4. Cumulative percentage of early func-
tions, MC-382 acceptance testing.

test conditions (ordinary temperatures and 30
grains of propellant) this distance of 525 ft
corresponds to a 0.7-sec flight time.

3. Upper limit for arming. The determina-
tion of the time or distance at which all fuzes
(excluding duds) will become armed is some-
what uncertain. The situation is complicated by
the effect of motor spin upon the action of the
switch. See Section 9.2.2 for discussion. An
estimate of an upper limit may be made from
the number of live units (total minus duds)
which functioned either early or on target in
acceptance testing. Blossom Point data show
that at least 98 per cent of the switches were
closed at 1,200 ft (1.4 sec, approximately). Re-
sults of the few arming tests, however, indicate
that when a reasonably satisfactory fin assem-
bly is used, the majority of the switches will be
armed after 1 sec of flight.11

Risk of Premature Function. The possibility
of the occurrence of a function before normal

370

ANALYSIS OF PERFORMANCE

time for closing of the arming switch is very firing at a target about 1,000 ft out, see Section
remote. In the assembling of hundreds of stand- 9.8 for requirements for acceptance) form the
ard units for acceptance testing and attendant greatest mass of data available concerning per-
experimental work, no switch was ever found formance of T-5 fuzes under fairly uniform fir-
to be in the armed position. Also in firing tests ing conditions. Although there were some differ-
no premature functions were ever observed on ences in conditions at the various proving
HE-loaded rounds. With inert-loaded rounds grounds, the setups were essentially the same,
using the highly sensitive spotting charges (see Details of procedure at the different locations
Chapter 8) the safety features of the powder may be found in Chapter 8.
train barrier do not apply. However, even in In Table 3 scores of acceptance tests are

Table 3. Acceptance testing results for T-5 fuzes.

Manu-

facturer

Proving

ground

Lot No.

No. fuzes
tested

Per cent

E L

Emerson

CC*

1-8, 10, 11

140

78.6

13.6

3.6

4.3

BPf

9 and 12-59

613

80.9

13.1

2.8

3.3

60-65, 71-97

338

83.4

11.2

1.8

3.6

Total

1,091

81.4

12.6

2.6

3.5

Friez

1-4

75.0

17.2

7.8

5-12 and 15

101

89.1

3.0

1.0

6.9

13-14, 18-261

120

86.7

8.3

5.0

Total

285

84.9

8.4

2.5

4.2

1-5

74.3

17.1

2.9

5.7

6-35

347

86.5

10.1

0.9

2.6

36-52, 55, 57-78

429

79.5

16.6

1.6

2.3

Total

846

81.9

13.9

1.4

2.7

Philco

1-11

166

73.5

14.5

3.6

8.4

12-58

570

83.5

11.9

0.9

3.7

59-77, 85-94, 98, 100,

380

84.2

10.5

2.4

2.9

102, 104-109

Total

1,116

82.3

11.8

1.8

4.1

Westinghouse

1-8

114

73.7

13.2

7.9

5.3

(Mansfield)

9-44

458

79.5

15.1

1.3

4.1

45-61, 65, 67-77

360

81.1

14.7

1.7

2.5

Total

932

79.4

14.7

2.3

3.6

Westinghouse

1-4

62.5

25.0

10.9

1.6

(Baltimore)

All

4,334

81.2

13.0

2.2

3.6

* Comcake Proving Ground, Fort Fisher, N. C.
t Blossom Point Proving Ground.
t Aberdeen Proving Ground, Aberdeen, Md.

these cases no fully verified premature func-
tions were reported. Occasionally (actually only
twice in many thousands of tests) observers
claimed to see the spotting charge operate as
the rocket left the launcher. Since visual rec-
ognition of the spotting charge during the
burning of the rocket propellant is extremely
difficult, the validity of even these rare obser-
vations is dubious.

Performance under Acceptance
Test Conditions

The results of acceptance testing (horizontal

listed according to manufacturer and proving
ground. Detailed analysis of the results ob-
tained at Corncake (Fort Fisher) and Blossom
Point may be found in reference 4. Although
these scores lead to a reasonably good estimate
of the overall performance for T-5 fuzes fired
under acceptance-testing conditions, two facts
should be pointed out: (1) individual scores
and variations therein cannot be taken at face
value as indicating corresponding variations in
manufacturing quality; nor (2) can exactly the
same performance as indicated by these accept-
ance results be expected of fuzes fired under

SECRET

FUZES FOR 4.5-IN. ARMY ROCKETS

371

conditions unlike those of acceptance work, i.e.,
high-angle, plane-to-plane.

It has been shown earlier that such factors
as motor type, propellant, and kind of trap may
very markedly affect early functioning. Still
other factors such as temperature and varying
distances from position of arming to target
must be taken into consideration. Since it is
not possible to separate these effects entirely,
lot-to-lot variation in performance must be
viewed with caution; specifically, for example,
the apparent improvement in scores of tests
conducted at Blossom Point over those done at
Corncake must not of necessity be taken as an
indication of improved manufacture, but rather
as the possible result of a combination of
many factors including perhaps even unknown
changes in test conditions.

Entirely apart from experimental test re-
sults, this view is amply supported by a study
of the acceptance test performance. For exam-
ple, Figure 5 shows that the February 1943
early functioning performance of fuzes of all
manufacturers was poor on Revere rockets, but
good on Budd rockets or Revere rockets fitted
with Budd fins. The strikingly uniform im-
provement in the subsequent performance of
all fuzes on Revere rockets is not accompanied
by any similar change on the Budd or modified
Revere rocket. No convincing explanation has
been found for the improvement on Revere
rockets.

In view of the results of later experimental
testing where attempts were made to control
increasingly larger numbers of variables (which
hitherto either had remained unnoticed or had
not appeared as relevant) the performance of
production fuzes appears to be satisfactorily
uniform. The overall score as given in Table 3
gives 81 per cent proper; it is safe to say that
with a satisfactory trap-ring-motor-propellant
combination a slightly higher score could now
be expected. (Much of the acceptance work
was done before high-angle testing showed the
importance of these three factors.)

Effect of Distance to Target
on Performance

Compared with those in actual combat use,
the distances between arming and target in ac-

ceptance testing were somewhat limited. This
means that in actual use, then, there would be
greater opportunity for the fuze to malfunction
before reaching the proper destination and, if
so, proper function scores would be lower.

Since the period for early functioning as de-
fined in Section 9.2.2 is about equal to the mini-
mum time taken for the SD feature of the T-5
to work, estimates of reliability for combat use
(for ranges longer than the acceptance test
range) can easily be made from results of high-
angle testing. The early functions remain clas-
sified as early and all other functioning rounds
become proper (see Section 9.2.2 for repre-
sentative scores). Figure 2, in Section 9.2.2,

Figure 5. Early function performance in ac-
ceptance testing of MC-382 rocket: Budd or
Revere with Budd fins (top) ; Revere, inert-
loaded or empty head (bottom). G General
Electric, E Emerson, F Friez, P Philco,

W Westinghouse.

shows time distribution of early functions for
various types of trap. When adjustment is made
for a 4 per cent dud score, the following per
cents proper obtain for rounds passing within

372

ANALYSIS OF PERFORMANCE

radius of action [ROA] of target for the indi-
cated flight times.

Flight time (sec.)

1.0

2.0

3.0
4.5

Estimated per cent proper
Worst trap Best trap

These values are based on the assumption
that performance for combat firing will be com-
parable to high-angle results. The early func-
tion scores in high-angle firing during the first
1.3 sec of flight and those for target testing
(about 1.3 sec to target) are comparable when
allowance is made for variations in motor con-
struction and propellant. This fact gives assur-
ance that the above estimates are fairly reliable.

Effect of Dispersion of Trajectories on
the Distribution of Bursts about a Target

Analysis by the Applied Mathematics Panel,
NDRC,41 of results with some thousand fuzes
tested on the mock-plane target (% scale of
B-25 bomber) at Blossom Point indicated, for
firing from astern, the following dependence of
target functioning upon distance of passage
from target axis (impact parameter).

Impact

Per cent of total rounds less
duds and earlies functioning

parameter (ft)

on target

100.0

99.9

99.3

95.0

79.7

50.8

21.5

5.5

0.8

100

0.1

The distribution of target functions for
rounds fired through ROA for acceptance re-
sults is shown in Table 4. In Figure 6 are shown
graphically the distributions for two values
of impact parameter. For empirical equations
to represent these distributions see reference 41.

Effectiveness in Plane-to-Plane Firing

A study was also made by the Applied Math-
ematics Panel to determine the probability that

a single 4.5-in. rocket fuzed with T-5 (when
fired from 1,000 yd directly astern) would dis-
able an enemy twin-engined bomber (Ju-88).
It was assumed that the rocket trajectories
have circular symmetry about the longitudinal
axis of the aircraft; specific dispersion data
used were from results at various proving
grounds. The value of fuze reliability used was
based on Blossom Point and Corncake data.
Estimates of damage by a projectile were based
on material presented in reference 40; these
estimates considered damage not only to the
engines but to various vulnerable portions of
the plane.

Calculations were made on two assumptions :
(1) that the plane could not return to base on
one engine, and (2) that the plane could return
to base on one engine only. The following re-
sults were obtained:

Standard deviation* Probability of disabling
of firing errors (ft) the aircraft

Aircraft assumed unable to return on one engine
25 0.207

50 0.106

75 0.057

Aircraft assumed able to return on one engine
25 0.143

50 0.066

75 0.035

* With a 50-ft firing error (standard deviation) the chance of a
direct hit is about one in a hundred.4i

Effectiveness in Plane-to-Ground Firing48

In a test to compare the effectiveness of VT-
fuzed and contact-fuzed rockets against person-
nel in slit trenches, 100 rounds of 4.5-in. rockets
(T-22, with T-23 fins), fuzed with T-5 were
fired from a plane over the effect field (Eglin
Field) described in Section 9.4.5. (Testing with
contact fuzes was discontinued after 18 rounds
fired — to ricochet — did not ricochet properly,
and gave an excessive number of low-order func-
tions.) Twenty each of the T-5’s were fired in
dive angles of 10, 20, 30, 40, and 50 degrees.

Table 5 shows the number of casualties
(scored as in Section 9.4.5) per burst at vari-
ous heights, for 4 degrees of shielding.

It is important to note that the significance
of “zero shielding/’ in this test, is somewhat
different from that appearing in some of the
literature on the effectiveness of air-burst pro-

FUZES FOR 4.5-IN. ARMY ROCKETS

373

jectiles. In this test the vulnerable area pre-
sented to any fragment moving in a horizontal,
or upward direction, is zero (unless the burst
occurs inside a trench) in the case of zero

sented to a burst occurring in the plane con-
taining the targets. The principal weakness of
the latter definition arises from the fact that
enemy troops would very rarely be distributed

Table 4. Distribution of functions in target firing of T-5 fuzes on Revere inert-loaded motors at Blossom
Point, March 1943 to March 1944.

Impact Parameter p — V#2 + z2 (ft)

3 1

8 2

3 2

8 3

3 3

8 4

3 4

8 5

3 58

Total

—5

—10

—15

—20

—25

100

202

180

335

£—35

130

261

—45

—50

—55

—60

—65

—70

Subtotal

182

481

228

972

Total

1 2

211

548

247

| 23

1 1,102

Table 5. Casualties as a function of burst height.

Burst

height

(ft)

12-in.
shielding
(conservative) *

Casualties

12-in.

shielding

per burst

6-in.

shielding

0-in.

shielding

0-5

0.6

0.8

1.0

1.9

6-15

1.8

2.3

3.0

5.9

16-30

1.8

3.1

3.7

6.6

31-50

2.4

3.7

4.3

6.1

51-80

1.4

2.2

2.7

3.9

81-125

1.2

1.5

2.5

4.2

* Counting only those bottom hits more than 6 in. from nearest edge of box (5 sq ft vulnerable area instead of 12 sq ft).

shielding. Although this definition is not be-
yond criticism, it is considered to be more prac-
tical than one alternative which considers that
the maximum possible vulnerable area is pre-

in a mathematically plane surface. For a fuller
discussion of this topic the reader is referred
to Section 9.4.5.

Figure 7 shows (1) mean burst height versus

SECRE'

374

ANALYSIS OF PERFORMANCE

dive angle, (2) casualties per burst versus dive indicate that the relative effectiveness of air
angle, (3) casualties per burst versus burst and ground bursts is not critically dependent
height. on the degree of shielding, and there is no rea-

Figure 6. Distribution of T-5 bursts along trajectories near fixed mock-plane target.

In actual combat, the casualties per burst son to expect that it would depend on the con-
would depend on the degree of concentration of centration of the enemy.

the enemy troops and on the shielding. The data At the optimum burst height, about eight

FUZES FOR 4.5-IN. ARMY ROCKETS

375

times as many casualties were obtained as with
ground bursts. On account of scatter in the
burst heights, and differences in the reflection
coefficient of various terrains, the optimum
height cannot be realized in combat. However,
the results show a rather wide range of burst
heights for which the relative advantage of the

DIVE ANGLE (DEGREES)

Code

BURST HEIGHT (FT)

Depth of
shielding
(in.)

Vulnerable
area
(sq ft)
12
12
12
5

Figure 7. Effectiveness of T-5 fuzed 4.5-in.
rockets for various degrees of shielding — plane-
to-ground firing: mean burst height as function
of dive angle (top left) ; casualties as function
of dive angle (top right) ; casualties as function
of burst height (bottom).

air burst over ground burst is larger and nearly
independent of the degree of shielding. This
useful range of burst heights is most closely
realized by firing in the steeper dives. Again
this is not at all critical, but dive angles in ex-
cess of 30 degrees are indicated (cf. footnote c
of Chapter 1).

9,2,4 Performance of T-6 Fuzes
Safety and Arming

General Characteristics of Arming Switches .
Mechanically the arming switch for the T-6 is
the same as that for the T-5. In addition further
delay is introduced by means of an electric
resistance-capacitance circuit. When mechani-
cal arming is completed, a switch is closed
which allows current to flow from the battery
through an arming resistor into the arming
condenser. The voltage on the condenser rises
until it is large enough so that a positive pulse
into the thyratron will cause it to fire and set
off the detonator. (There is a region of time
just before the condenser is charged sufficiently
to complete the arming cycle during which a
pulse on the input to the thyratron will cause it
to become conducting. This removes a portion
of the charge accumulated, without firing the
detonator. This phenomenon, called “dumping”
(cf. Section 3.3.6), occurs only if the fuze re-
ceives a firing signal sometime during the in-
terval when the condenser has voltage enough to
ignite the thyratron but does not contain
energy enough to fire the detonator. If such an
accidental “dumping” signal occurs, the circuit
automatically recovers and arms at a time about
20 per cent longer than normal. This phenom-
enon does not cause serious trouble under ordi-
nary circumstances.

Arming Time and Distance. Direct measure-
ment of arming time cannot be made for
switches incorporating an RC delay (see Sec-
tion 8.3.7). However, from laboratory determi-
nations of values of the various electric com-
ponents, along with measured times to mechan-
ical arming, satisfactory predictions of arming
times for the T-6 can be made.23' 24

The validity of such predictions is substanti-
ated by field tests of Navy rocket fuzes (see
Section 9.3.2) in which fuzes were “pulsed”
during flight by a transmitter located on the
firing range. Such tests do not determine the
arming time of an individual fuze, but they do
give an experimental lower limit on the fraction
of fuzes fully armed at any given time.

Figure 8 shows the per cent of fuzes armed
as a function of time and of horizontal range
for rounds fired on Revere 4.5-in. rocket (V0 =

376

ANALYSIS OF PERFORMANCE

840 fps). It will be seen that no fuze can func-
tion before a horizontal range of 800 yd and
95 per cent will be armed at 1,650 yd. (If the
projectile passes within 150 ft of crests or other
suitable targets before arming is complete,
“dumping,” as mentioned above, may occur.
Under these conditions the percentage of

Figure 8. Cumulative per cent of T-6 fuzes
armed as function of flight time and horizontal
range.

fuzes armed may be slightly reduced at ranges
up to 2,000 yd.)

Reliability

No acceptance testing, as such, was done with
the T-6 fuze. Since except for the arming switch,
T-5 and T-6 are identical, estimates of reliabil-
ity may be made from results of experimental
high-angle testing of the T-5. An intensive
study32 of middle functions, random functions
occurring after 5 sec, among some 1,600 rounds
yields the following estimates of performance
as a function of flight time.

Flight time

Per cent proper funct

(sec)

Mfr A

Mfr B

The above percentages are based on rounds
fired on 4.5-in. rockets with reasonably satis-
factory fin assemblies. For discussion of the
effect of fins on malfunctioning of the T-6, see
Section 9.2.2.

Effectiveness

In tests of the relative effectiveness of 4.5-in.
Army rockets using T-6 fuzes, PD M-4 fuzes
set for ricochet air bursts, and PD M-4 fuzes set
for superquick action, 260 rounds were fired
over an effect field at Fort Bragg, North Caro-
lina.50 The field contained lx6-ft boards spaced
5 yd apart, laterally and longitudinally. The
boards were laid in shallow trenches with top
surfaces 1 in. below ground level. On each
round which burst on the effect field, the num-
ber of boards hit by at least one fragment which
penetrated at least % in. into the wood was
counted.

The results are given in Table 6.

Table 6. Comparative effectiveness of T-6 fuze
and PD M-4 fuzes set for ricochet air burst and
for superquick action.

Total
No. of
rounds
fired

No. of
rounds

effect

field

Average
height
of burst
(ft)

Average
No. of
targets
hit
(per
round)

T-6

21.2

PD M-4, ricochet air

burst

16.8

PD M-4, superquick

action

4.4

These results, where heights are visual esti-
mates, are quantitatively somewhat different
from those obtained in T-5 testing at Eglin
Field, where heights were photographically de-
termined. (See Figure 7.) In the test of the T-5,
greater effectiveness was observed at a burst
height of 17 ft than at one of 60 ft. The Fort
Bragg results do agree with those obtained at
Eglin Field, however, in that they indicate a
fourfold or fivefold advantage, over a contact
burst, of an air burst occurring over a consid-
erable range of heights.

NAVY ROCKET FUZES

377

9 3 NAVY ROCKET FUZES

931 General

This section concerns the performance of the
VT fuzes which are intended primarily for use
on rockets as follows :

Fuze

Navyord. Army ord.

designation designation Use Rockets

Mk-172 T-2004 Plane to ground AR 5.0

Model 0

Mk-171 T-30 Plane to plane HVAR

Model 0

Throughout the section the following rocket
designations, established by the California In-
stitute of Technology [CIT], are used for con-
venience. The aircraft rocket [AR] 5.0 is used
to designate the 5.0-in. Mk-1 shell with the
3.25-in. Mk-7 motor. The high-velocity aircraft
rocket [HVAR] refers to the same shell, or the
5.0-in. Mk-5 shell, used with the 5.0-in. Mk-1
motor. Considerable testing was performed
with the AR 3.5, denoting the 3.5-in. Mk-5 or
Mk-3 (16 lb) shell and the 3.25-in. Mk-7 motor
combination.

The same scoring terminology and methods
that are used for the Army rocket fuzes (see
Section 9.2) are applied to the Navy rocket
fuzes.

The mass production model of the T-2004 fuze
was subjected to two stages of acceptance test-
ing. The first stage, the “metal parts” accept-
ance test, was applied to a sample from each
manufacturers’ lot of approximately 1,000
“metal parts assemblies,” which constituted a
“metal parts lot.” A spotting charge was used,
and the rockets were not loaded with high ex-
plosive. Following acceptance, these assemblies
were loaded with the few additional explosive
components necessary to make complete fuzes,
and the lots were usually combined into much
larger lots known as “ammunition lots.” The
second stage was the ammunition lot acceptance
test, applied to a sample from each ammunition
lot, to check on the safety and reliability of the
complete fuzes that were subsequently shipped
to the using Services. A discussion of compo-
sition of ammunition lots and the relation be-
tween their performance and that of the metal

parts lots is given in Section 9.4.3. Procedures
for acceptance testing are outlined in an appen-
dix to this chapter.

The large difference in overall sensitivity
makes it desirable to treat the T-30 and T-2004
separately except for their arming character-
istics. The latter are identical mechanically but
differ in the amount of RC delay.

The relation between early functioning and
afterburning is discussed under the T-30, since
the problem is of much greater importance
with the fuze that has the greater sensitivity
and shorter arming time. Also, the afterburn-
ing of the HVAR is much more serious than
that of the AR.

Section 9.2.2 should be consulted for a dis-
cussion of basic considerations relating to after-
burning as well as for the background provided
by experience with the T-5 Army rocket fuze.
Conclusions concerning afterburning and early
functioning of VT rocket fuzes in general are
given at the end of Section 9.3.3, with partic-
ular reference to the T-30 and the Navy rockets.

932 Safety and Arming

General

The arming mechanism of the VT fuzes for
Navy rockets is so designed that mechanical
arming occurs when acceleration of the rocket
ceases. Complete arming is delayed somewhat
further by the use of an RC circuit.

Mechanical arming tests should provide data
that are in general agreement with the burning
times and distances of the rockets. No exact
comparison is practical, however, since burning
does not cease abruptly, and mechanical arming
occurs at some time during the final “tapering
off” of the burning.

The values of the RC arming network follow.

Fuze

(megohm)

(mf)

T-2004

1.5

1.0

1.50

T-30

0.82

0.90

0.74

The average RC delay should be approxi-
mately equal to the product RC in seconds (see
Section 3.3.6). For the sake of completeness
tests with fuzes having other values for R and
C are included in the following analysis.

378

ANALYSIS OF PERFORMANCE

In addition to tests of arming performance,
there are summarized the results of a safety
test. This test, which was performed primarily
as a check on the safety of the value adopted
for arming distance, is of particular importance
because it was conducted under conditions sim-
ulating rather closely those of Service use of
the fuzes.

Mechanical Arming Performance

The most accurate data on mechanical arm-
ing performance are probably those obtained
in the experimental tests, summarized in Table
7,5G in which most of the arming distances were

The relatively large arming distance observed
with the AR 3.5 is probably due to the lower
efficiency of the propeller at the high speed of
this rocket, which is about the same as the
velocity of the HVAR.

The best estimate of spread in mechanical
arming distances (from Table 7) is given in
Table 9.

From an inspection of Table 7 it is evident
that there is a lower temperature limit, in the
neighborhood of —20 F, for the reliable oper-
ation of the mechanical arming device with the
AR 5.0. At —20 F, about half of the arming
mechanisms failed to function. This limit arises

Table 7. Results of experimental FOMA tests.

SD of

Mean powder

Standard

individual

No. of

Score

temp

Mean arming

error of

arming

units

FOMA-D-L

(degrees F)

distance (ft)

mean (ft)

distance (ft)

Bowen and GE T-30 units on AR 5.0

9-1-0

461

10-9-1

—20

484

9-1-0

—10

476

9-1-0

465

10-0-0

439

Philco T-200Jf units

on AR 3.5

14-1-1

596

Philco T-2004 units

on AR 5.0

14-1-0

445*

, .

9-1-0

415f

* No photographic data of arming distances are available for this test. The average arming distance (445) was computed by assuming
the average velocity in this test to be the same as the average velocity of similar rockets of other tests when fired under similar test con-
ditions. The arming distance was then computed from the average speed and the observed arming time (1.086 sec).

t This figure is based on the photographic data of only 3 units.

determined photographically. Arming times
(stopwatch measurements) ranged from about
1.0 sec at the higher temperatures to 1.6 sec at
the lowest. The distances given in Table 8,
from ammunition lot acceptance tests, were
calculated from stopwatch measurements of
arming time. Making due allowance for timing
errors, the values agree reasonably well with
those in Table 7.

For an analytical comparison of temperature
effects on arming distance with temperature
effects on burning distance, reference 37 should
be consulted. Here it is sufficient to note from
Table 7 that the effect of temperature on arm-
ing distance is practically negligible through-
out a rather wide temperature range.

from the fact that a certain minimum accelera-
tion is required for arming of the fuzes. Tests by
Army Ordnance indicate an upper temperature
limit in the neighborhood of 110 F. An upper
limit arises from the fact that the propeller
must make a minimum of approximately 100
turns before acceleration falls below a certain
value. For a complete discussion of the me-
chanics of the arming device see Chapters 4
and 5.

Results of Pulsing Tests to Obtain Total
Arming Distances of T-30 Fuzes

Data are provided by a number of tests de-
signed to determine the spread in arming times
and arming distances of T-30 fuzes31 on the

NAVY ROCKET FUZES

379

AR 3.5. In these tests each fuze was “pulsed”
by a transmitter at a certain point in its flight
to determine whether or not it was armed. The
experimental technique is covered in Section 8.3.
The tests are very difficult to perform and the
data are, therefore, rather limited.

Table 8. Results of Philco acceptance FOMA tests
of T-2004 fuzes on the AR 5 rocket.

PA lot

No.

PA-315

No.

of Score

units FOMA-D-L-I

Powder

tempera-

ture

(degrees

Mean

arming

distance

(ft)

10-0-0-0

72 )

458

7-0-1-2

72 \

9-0-0-1

542

8-2-0-0

453

10-0-0-0

404

10-0-0-0

554

7-1-0-0

428

Table 9.

Mechanical arming spread of T-30 and

T-2004 units on AR 5.

Amount

Pooled

Amount

beyond

Total

estimate

short of

the mean

spread

the mean

for

1% to

standard

for 1%

95%

deviation

armed

Arming

distance

49 ft

129 ft

91 ft

220 ft

From such pulse test data it is possible to
calculate a probability distribution of total arm-
ing distances. The results are summarized in
Table 10. No pulsing tests performed with other
rockets yielded sufficient data for probability
distributions.

Table 10. Results of pulsing tests of T-30 units on
AR 3.5, 1.15-mf firing condenser.

Delay

Arming time

Arming distance

resistor

1% median 95%

median

95%

(megohms)

(sec)

(ft)

0.51

1.09 1.36 1.58

740

1,060

1,310

0.75 '

1.26 1.53 1.76

940

1,240

1,480

The analytical comparison of the data of
Table 10 with engineering prediction is too
complex for presentation here. The difficulty of
the tests was increased by lack of accurate bal-
listic data for VT-fuzed rockets, and reference
31 should be consulted for an adequate treat-
ment of the data. Essentially, the analysis

showed that there was satisfactory agreement
between prediction and observation if it was
assumed that one “dumping” cycle occurred in
most of the fuzes before complete arming (see
Section 3.3.6 for description of dumping).

Calculated Percentage Points for
Total Arming

It is indicated in the preceding section that
minimum safe arming distances can be pre-
dicted from data on mechanical arming per-
formance together with the engineering theory
of RC arming. In order to make conservative
predictions, it is desirable to assume that
“dumping” does not occur. Maximum arming
distances calculated on this basis are likely to
be underestimates. However, since the number
of “dumping” cycles that occur is likely to de-
pend considerably on any condition, such as
temperature, that affects afterburning, it ap-
pears desirable to calculate the maximum arm-
ing distance on the same basis as the minimum.
Calculated values are given in Table 11 for sev-

Table 11. Mechanical and total arming distances

of rocket fuzes, 1.15-mf firing condenser.

Delay resistor Arming distance (ft)

(megohms) 1% Median 95%

Mech. arming

AR 3.5
350

600

740

0.51

740

1,060

1,300

0.75

930

1,240

1,480

1.50

1,460

1,760

1,980

Mech. arming

AR 5.0
330

460

550

0.51

450

790

1,020

0.75

650

970

1,200

1.50

1,180

1,470

1,700

Mech. arming

HVAR

420

650

810

0.51

940

1,140

1,530

0.75

1,160

1,360

1,760

1.50

1,830

2,080

2,630

eral values of arming delay resistance. Rocket
ballistic tables38 were used in making estimates
for the HVAR. Table 11 is probably most re-
liable when used in connection with the T-2004,
in which “dumping” is less likely to occur. In
connection with the T-30, especially on the
HVAR, it should be remembered that the actual
maximum (or 95 per cent) arming distance is
likely to be in excess of tabulated values, on
account of the “dumping” phenomenon.

SECRET

380

ANALYSIS OF PERFORMANCE

Tests of Safety

Safety tests were conducted by the Navy Bu-
reau of Ordnance at Inyokern42’ 43 to determine
the fragmentation effect, in cases of early func-
tioning. In these tests a TDR drone was modi-
fied to fire HE-loaded AR 5’s, fuzed with T-30
or T-2004 fuzes, wired to fire on mechanical
arming. The rockets were fired from under the
wings of the drone, when the aircraft reached
maximum airspeed in a maximum dive angle
of about 10 degrees. The results of 27 rounds,
all of which functioned on mechanical arming,
showed no hits on the drone. Since these fuzes
had no normal RC delay, it was concluded that
rearward fragmentation damage to the firing
plane was a very remote possibility.

Performance of T-30 Fuzes

Practically all the data on T-30 performance
were obtained with pilot production models.
Some of the early testing was done with modi-
fied bomb fuzes. Two peak amplification fre-
quencies, approximately 100 and 70 c, were
tried during pilot production, and there were
a number of other variations, including changes
in RC delay resistor, generator shaft couplings,
and thrust bearings.

Because of difficulties with dispersion in fir-
ing the Navy rockets from a fixed launcher at
a mock-plane target (see Chapter 8) most of
the testing was performed by firing at high
angle or from a plane for function on approach
to water. The relation between the scores ob-
tained in the two types of tests has already
been discussed in Section 9.2.

Afterburning and Early Functioning17

General. Before discussing the performance
of the T-30 in the conventional types of tests
just mentioned, it is desirable to give some at-
tention to the problem of early functioning and
afterburning. A VT fuze on Navy rockets has
to be armed at some time subsequent to the
main burning time to avoid malfunctioning due
to afterburning. This fact puts a serious limi-
tation on the tactical effectiveness of the VT
fuze on these vehicles.

Experience with the T-5 Army rocket fuze

had shown the importance of afterburning (see
Section 9.2). In field tests of developmental
models of T-30 on Navy rockets it was noted
that there was a great deal of afterburning
from the motors, and the poor performance of
the fuzes was attributed to this afterburning.
These early fuzes had no RC arming delay.

A cooperative investigation between the
Navy, Division 3 and Division 4, NDRC, was
started at the Naval Ordnance Test Station
[NOTS], Inyokern, in order to find means for
reducing the effects of afterburning on HVAR.

The Propellant Grain and Its Burning Char-
acteristics. The cruciform grain of Ballistite
used in the HVAR has a length of 39.5 in. and
an outside diameter of 4.20 to 4.26 in. After
ignition the burning is maintained at nearly
constant rate by means of inhibitors. The burn-
ing progresses until the surface of the grain
has decreased so greatly that the resulting pres-
sure will not support the primary burning. A
core of unburned Ballistite remains at the end
of the main burning. After the main burning,
the core continues to receive heat from the
motor wall and nozzles; its temperature is
raised, and secondary burning is initiated. The
secondary burning continues until the core is
either consumed or becomes small enough to be
ejected through the nozzle.

The rate of secondary burning is so low that
negligible contribution is made to the forward
thrust of the rocket. The core is, therefore, use-
less to the rocket and far more useless to the
fuze, since afterburning causes malfunctions.

Static Tests in an Air stream. Static tests at
Alleghany Ballistics Laboratory and at Inyo-
kern, conducted by placing the rocket in a
stream of air to simulate some of the conditions
of flight, showed definite correlation of fuze
pulses with afterburning. Afterburning of the
Ballistite caused pulses which were several
times as strong as the pulses necessary to
trigger the fuzes.

Since afterburning depends on the presence
of the core, it was decided to eliminate a large
portion of the core by extruding the Ballistite
grain with an axial perforation. The presence
of the perforation suggested the possibility of
filling the void with some substance which
might be beneficial in overcoming the after-

SECRET

NAVY ROCKET FUZES

381

burning. Sand, table salt, hypo, sal soda, borax,
alum, and Epsom salts were tried. Empty per-
forations were also tried. Comparisons of these
loadings were made with standard grains.
Hypo, alum, and Epsom salts proved to be the
best. The others were less effective but superior
to standard grains. It may be that evaporation
of water of crystallization in some of these ma-
terials might cool the gases to such an extent
that afterburning would not be started. It was
found to be equally effective to place quarter-
pound bags of hypo, wrapped in cloth, ahead of
the igniter. As a practical measure, hypo may
be unsatisfactory, because it melts at 120 F
(this temperature could be easily exceeded in
motors exposed to summer sun), and even if it
did not melt, the crystals would yield water
vapor, which would be absorbed by the Bal-
listite, where it might cause trouble. For these
reasons, alum or Epsom salts would be better,
because each of these contains as much or more
water of crystallization and yet has a lower
partial pressure of water vapor. Common salt
and sand, which have no water of crystalliza-
tion, give much better results than standard
rounds, but inferior to those containing water
of crystallization. The fact that hypo, alum, and
Epsom salts, which gave the best performance,
are sulfates suggests that the presence of sul-
fur may be an important factor.

Results with hypo showed no afterburn- *
ing and no pulses in any of 13 static experi-
ments.

Ground-Launched HVAR Tests. The fuze had
performed well on the HVAR, fired statically
in the airstream, with hypo in the perforation
in the grain. However, when such rounds were
fired from a ground launcher, the addition of
hypo bags impaired significantly fuze perform-
ance in flight. The fact that hypo bags had
eliminated the afterburning in the static tests,
yet increased the early functioning in the flight
tests, is one of the paradoxes in the afterburn-
ing program.

The HVAR rockets, modified to have single
nozzles in place of standard multiple nozzles,
were fired but failed to indicate a significant
difference in performance from standard
HVAR. It was thought that the single nozzle
would permit ejection of the core at the end of

primary burning, as is often observed to be
true of AR rockets.

Plane Firing with HVAR. The final appraisal
of T-30 performance must come from plane
launchings, since this type of testing is nearest
to tactical conditions. Fuzes must be ready to
function after the rocket is at a short yet safe
distance from the firing plane. Effectiveness
will be limited by the proportion of duds and
random functions, the lateness of arming, and
rocket dispersion.

The results from rounds fired from a plane in
a dive at various ranges indicated an average
time delay of 0.3 sec due to dumping of the
firing condenser caused by afterburning. This
extra time is much shorter than was expected
from the static firings and shows that static
firings cannot be relied upon to indicate the
performance of plane-fired rounds.

The variation in fuze performance under
different test conditions is further accentuated
by comparing early-function scores of ground-
launched and plane-launched rounds. A test at
Inyokern of T-30 on standard HVAR, fired at
a slant range of 2,500 yd from a plane in a 30-
degree dive flying at 200 mph, yielded 5 duds,
58 propers, and 27 earlies (32 per cent early-
function score). The early functions centered
at 1.85 sec, which is only slightly greater than
the arming time. These results may be com-
pared with the data in Table 12 for ground-
launched HVAR, which show a 17 per cent
early function score.

Conclusions. The main conclusions obtained
from the study of malfunctioning of VT fuzes
and afterburning of rocket motors are as fol-
lows: Many correlations have been made of
afterburning and VT fuze malfunctioning, but
as yet very little has been definitely proved
about the fundamental causes of afterburning.
Static experiments, using a 110-fps airstream
past the rocket nozzle, correlate the afterburn-
ing with pulses on the fuzes and indicate that
a large portion of malfunctioning on the HVAR
would be due to afterburning. Hypo, which de-
creased the afterburning in static firings,
caused an increase of early functions on
ground-launched rounds. On the other hand,
standard HVAR rockets launched from a plane
produced a larger proportion of early func-

SECRET

382

ANALYSIS OF PERFORMANCE

tions. This suggests a real difference in per-
formance, but it is difficult to explain why an
increase in the airspeed of a rocket from a
supersonic velocity to a higher supersonic ve-
locity would increase the proportion of early
functions. Therefore, firing fuzed rockets from
airplanes, though a more difficult procedure,
seems to be the only reliable procedure to use in
any further study of the phenomenon that may
be undertaken.

Performance in Firing for Function
On Approach to Water

The performance of T-30 fuzes fired from a
ground launcher for function on approach to
a water surface is summarized in Table 12.
The summary for fuzes fired from a plane for
function on approach to water is given in Table
13. Burst heights are included as a matter of
general interest in Table 13, although they are
of secondary significance in connection with the
intended air-to-air application of the fuze. Re-
sults obtained with fuzes with different ampli-
fier characteristics (indicated by the frequency
of maximum gain [PkAF] ) are pooled in
Table 12, since there was no evidence of any
effect of this difference on the scores. Burst
heights are affected by PkAF and are, there-
fore, not given in this table.

It should be noted that the velocity of the
light AR 3.5 (AR 3.5 with the special 4-lb
shell) is about the same as that of the HVAR,
while the regular AR 3.5 and AR 5 are pro-
gressively slower (in the order named). Ac-
cording to experienced observers, afterburn-
ing is most pronounced ^vith respect to both in-
tensity and duration in the HVAR.

In examining the high-angle test results in
Table 12, it will be noted that although the dud
scores are a little high, they are not excessively
so in comparison with most production model
VT fuzes. For equal flight times (quadrant ele-
vations), middle functioning is markedly
greater on the HVAR than on the light AR 3.5.
A rational explanation is provided by the pro-
longed afterburning of the HVAR Early func-
tioning is uniformly about 20 per cent on the
high-speed rockets, and significantly less on the
regular AR 3.5. The excess on the light AR 3.5
is most plausibly associated with the greater

intensity of mechanical vibration of the rocket
that is to be expected at the higher speed. If the
difference were accounted to electric disturb-
ances caused by the higher generator speed, a
still greater excess of early functions might be
expected on the HVAR, where afterburning is
so prominent.

Since 5 sec may be taken as a maximum use-
ful flight time for plane-to-plane firing, it is
reasonable to combine the middle-function

Table 12. Performance in high-angle firing.

Reference

number

Number

fired

Per cent

P E M

Quadrant

elevation

(degrees)

252

AR 3.5

83 9

122

HVAR

55 20

112

AR 3.5*

64 19

216

AR 3.5*

55 21

* AR 3.5 with a special 4-lb shell for testing VT fuzes.

scores with the proper-function scores. The
data in Table 12 then give rather uniform
proper-function scores of 73 to 77 per cent on
the high-speed rockets. A reliability of about
75 per cent is thus indicated for rounds well
within the radius of action of an airplane target
at extreme range.

Special comment is needed on the results of
plane-to-water firing given in Table 13, and it
may as well be admitted at the outset that most
of the data would have been omitted had there
been anything more reliable to offer for study.
The first three lines of data in this table repre-
sent results obtained with six missions of eight
rounds each. Subsequent high-angle firing, and
the test summarized in the last entry in the
table, gave a very strong indication that the
excessive dud scores were due to the use of
Fahnstock clips on the arming wires (some of
which were reported broken after these tests).
Unfortunately, carrier observations were not
made in the initial test, so that it was not pos-
sible to localize sharply the source of the
trouble. The differences between the first three
dud scores may be due to the differences in the

SECRET

NAVY ROCKET FUZES

383

Table 13. Performance of T-30 in plane-to-water
firing.64* Flight time, 4 to 5 seconds.

Number

Per cent

Mean

burst

Standard

error

PkAF

fired

P E D

height

(ft)

mean

(ft)

(c)

AR 3.5 Dive angle 30-50° ; dive speed 235-250 mph

16 75 0 25 104 5 77

AR 3.5* Dive angle 30-50° ; dive speed 235-250 mph

16 100 0 0 95 4 77

AR 5.0 Dive angle 30-50° ; dive speed 235-250 mph

16 50 6 44 102 6 77

AR 5.0 Dive angle 40° ; dive speed 320 mph

23 91 9 0 325 17 100

* AR 3.5 with a special 4-lb shell for testing VT fuzes.

initial accelerations of the three rockets. How-
ever, the differences cannot be regarded as
having much statistical significance in view of
the fact that there were only six missions and
in view of the likelihood that the arming wire
installations were probably fairly uniform for
each mission.

If it were not for the possibility that some of
the duds were due to the “dumping” phenome-
non, one might adjust the dud scores to be equal
to the average of the values given in Table 12.
However, the only safe conclusion is that there
is no inconsistency between the plane-to-water
firing results, and the high-angle firing results
when due allowance is made for duds caused by
the use of Fahnstock clips on arming wires in
some of the plane-firing tests.

Performance in Firing at a
Fixed Mock-Plane Target

There were a number of mock-plane tests at
Blossom Point, using various rockets and vari-
ations of fuze design. Many of the tests were
made, using modified T-50 fuzes27 before T-30
models had been constructed. Shortly after a
satisfactory design for the T-30 had been de-
veloped, emphasis was shifted to the T-2004 for
air-to-ground firing; hence, only a small num-
ber of rounds were fired against the mock plane
with the final design of T-30.

The results of T-30 target tests are given in
Table 14. In order to indicate the radius of

action there are included the observed impact
parameters p of trajectories defined as

p max = distance in feet of the farthest tra-
jectory from the target for target functions.

p min = distance in feet of closest trajectory
to the target for any fuze that functioned far
beyond the target.

Table 14. Performance in firing at a fixed mock-
plane target.

Pro-

jectile

Number

fired

Per cent
E L

Max

Min

PkAF

AR 5.0

100

AR 3.5*

100

AR 3.5*

101

lOlf

AR 5.0

89f

* AR 3.5 with a special 4-lb shell for testing VT fuzes,
f Two fuzes which passed the target at 33 and 34 ft respectively
and later functioned on approach have been omitted from considera-
tion. These were either unusually insensitive or were not armed at
the target because of possible dumping of the firing condenser.

Spot charts showing the position and dis-
tance of closest approach (impact parameter)
of all target and passage functions may be
found in reference 29.

Fuzes with amplifiers peaked at 100 c appear
to have a radius of action of approximately 70
ft; fuzes with peak frequencies of 77 c show a
radius of action of approximately 90 ft. Except
for the two rounds which passed the target at
33 and 34 ft, the bulk of the live rounds which
passed the target without firing began at 70 or
90 ft respectively. The results for fuzes with
trajectories within the radius of action are as
follows :

PkAF , 100 c; radius of action, 70 ft
N P E L D </cP

16 15 0 0 1 94

PkAF, 77 c; radius of action, 90 ft
N P E L D %P

29 22 0 3 4 76

The almost complete absence of early func-
tions, shown above, is to be attributed to the
short distance from projector to target (1,200
ft).

Performance in Plane-to-Drone Firing

Firings against a drone were conducted at
the Naval Ordnance Test Station at Inyokern.
The T-30 fuzes on AR 5.0 and AR 3.5 were

fSECHET

384

ANALYSIS OF PERFORMANCE

ripple fired from a torpedo bomber [TBM]
equipped with four zero-length rails, against a
TDR drone.42 Spotting charge-loaded as well
as HE-loaded rounds were used. Precise meas-
urement of burst positions was impossible be-
cause of the test conditions. The firing was done
from about 375 yd astern of the drone at speeds
of 160 knots (pursuit) and 95 knots (target
plane). It should be noted that, although the
maximum speed and the reflection properties of
the drones employed are somewhat different
from those of combat planes, the test condi-
tions were much more like those of combat than
any other tests performed with this fuze. Re-
sults of the Inyokern tests are given in Table
15.

Table 15. Performance in plane-to-drone firing.

AR 3.5

AR 5.0

spotting

AR 5.0

charge

HE-loaded

Early (%)

Passage without function
(%)

Proper (%)

Number fired

Mean distance of propers
from target (ft)

Several drones were destroyed by the HE-
loaded rounds. It was obviously impractical to
destroy enough drones to obtain a reliable

not be calculated by the usual means. However,
the conditions of the test were such that the
radius of action would be expected to be
roughly 50 per cent greater than the mean dis-
tance of proper functions. Considering the
assumptions involved in estimating the radius
of action in this way, the 70-ft value so ob-
tained is in satisfactory agreement with the
estimate of 90 ft from the fixed target tests. A
comparison of proper function scores cannot be
made, since the number of rounds within the
radius of action of the drone is unknown.

934 Performance of T-2004 Fuzes

The T-2004 is designed for use on rockets
fired from a plane against targets on land or
water. It has, therefore, been possible to test
this fuze under the conditions of its tactical use.
However, because of the greater convenience,
the larger part of the proof testing has been
done by firing from a stationary launcher for
function on approach to water. Results of ex-
perimental high-angle firing tests are summar-
ized in Table 16. Results of experimental plane-
to-surface firing tests are summarized in Table
17. Most of the tests in Tables 16 and 17 in-
volved pilot-production models with a number
of variations in arming delay resistance, gen-
erator shaft couplers, and thrust bearing.

Table 16. Performance in experimental high-angle firing. Target factor: 81.

Standard

Approximate

Mean

error

Reference

flight time

Number

Per cent

burst height

mean

No.

(sec)

fired

E M

(ft)

288

AR 3.5, quadrant elevation 30°
92 4 1

0.4

AR 5.0, quadrant elevation 30°
85 0 0

1.1

HVAR, quadrant elevation 30°
64 20 4

1.9

AR 3.5, quadrant elevation 70°
92 4 0

1.2

measure of the probability of so doing. The
weighted overall mean distance of the proper
functions is 46 ft. Since the passage distances
could not be measured, the radius of action can-

Acceptance tests of the production model
T-2004 fuzes were conducted on the AR 3.5
according to the procedure outlined in Section
9.8. Results are summarized in Tables 18 and

NAVY ROCKET FUZES

385

20. In all cases where the target factor (reflec-
tion coefficient in per cent) is given as 81, the
target was a water surface. Acceptance testing
was done at Aberdeen (target factor 81) and
Jefferson Proving Grounds (target factor 65).
Target factors for firing against ground at
Dahlgren and Inyokern are unknown.

so rare that the available data are insufficient
to give a reliable time distribution. The accept-
ance test data do show, however, that the inci-
dence of functions from arming up to 5 sec is
roughly ten times as great as it is during any
later equal period. The 5-sec limit may, there-
fore, be regarded as a useful one, even though

Table 17. Performance in experimental plane-to-surface firing.43- 64b

Approximate
flight time
(sec)

Number

fired

Per cent
E

Mean

burst height
(ft)

Standard
error mean
(ft)

Target

factor

HVAR, 30 ° dive, plane speed 320 mph

Land

4.7

U.5" T-87, 30° dive, plane speed 250 mph

1.2

AR 5.0 , 30° dive, plane speed 310-3 40 mph

1.4

AR 5.0, 40° dive, plane speed 345 mph

Land

Inspection of the tabulated flight times
shows that the time available for middle func-
tioning was very limited in the plane-firing
tests, and no such functions were recorded in
the tests of Table 17. In several of the tests
summarized in this table, there was a very lim-

its significance in relation to afterburning is
not well defined.

It should be noted that the classification,
“late” (functions below 10 ft), used in the ac-
ceptance tests, is a carry-over from the accept-
ance testing of bomb fuzes (see Section 9.4).

Table 18. Performance of T-2004 in metal parts acceptance tests.

Lot-

group

No.

Approximate
flight time
(sec)

Number

fired

Per cent

P E M L D

Mean

burst height
(ft)

Standard
error mean
(ft)

Target

factor

817

Quadrant elevation 30°

90 1 1 2 6

0.4

100

30° dive, plane speed 250 mph
88 2 2 1 7

1.8

877

Quadrant elevation 20°

95 2 1 0 2

0.3

Quadrant elevation 42°

92 6 0 0 2

0.6

6.5

20° dive, plane speed 290 mph
98 0 0 0 2

0.7

* See Table 19 for identification of lots.

ited opportunity even for early functions. In
this connection it should be noted that the
5-sec time limit used in the classification of
early functions of the Navy rocket fuzes is a
carry-over from the testing of the T-5 Army
rocket fuze. Malfunctioning of the T-2004 is

This classification was not used in the experi-
mental tests in which either the Army or Navy
rocket fuzes were fired for function on ap-
proach to a ground or water target. The accept-
ance usage was based on considerations of mili-
tary utility rather than on the existence of any

386

ANALYSIS OF PERFORMANCE

discontinuity in the function height distribu-
tion (see Section 9.1.3).

Proper functioning performance on the AR
3.5 ranges from 88 to 98 per cent in the tables.
Information on performance on the AR 5.0 is
limited to pilot production models, and per-
formance is not quite as good, averaging about
90 per cent proper functions. The deficiency is
due mainly to the rather high incidence of duds.
This may easily be attributable to difficulty

Table 19. Composition of lot groups in Table 18.

Lot

group

No. Metal parts lots

1 1001-1060 excluding lots ending in the
digit 8

2 1008, 1018, 1028, 1038, 1048, 1058

3 1061-1066, 1072A, 1074A, 1075-1097, 1099,

1100-1107, 1109-1117, 1119-1122

4 1067, 1069, 1074

5 1068, 1078, 1088, 1098, 1108

with detonator contact spring adjustments.
There is evidence of a progressive reduction in
dud scores throughout production, as shown in
Table 18. This parallels experience with bomb
fuze production as shown in Section 9.4.3,
where the phenomenon is discussed more fully.
There is no known reason why performance
should be better on the AR 3.5 than on the
AR 5.0, and it is probably quite safe to take
the acceptance data as representing the relia-
bility of the fuze on both rockets.

Table 20. Performance of T-2004 in ammunition
lot acceptance tests. Quadrant elevation: 20° in
most cases; target factor: 81; lots: 315-2 through
315-20.

Approxi-
mate Mean Standard

flight burst error

time Number Per cent height mean

(sec) fired P E M L D (ft) (ft)

19 177 94 2 1 1 2 32 1.1

Scoring performance on the HVAR is con-
servatively represented by the data in Table 16,
since a short RC arming delay (0.75 megohm,
1.0 mf) was used in the test. The same delay
was used in the HVAR tests given in Table 17.

These results have little significance except for
very short flights.

It is not possible to make a satisfactory com-
parison of observed with predicted burst
heights on account of the absence of reliable
terminal ballistic data on VT-fuzed Navy rock-
ets. Approximate calculations indicate that, as
in the case of bomb fuze burst heights, there
are some inconsistencies in the burst heights
recorded for the different test conditions in
Table 18. None of the discrepancies is serious,
however, and a fuller discussion is unwarranted
here.

Particularly interesting data on burst
heights were obtained in the Naval Ordnance
Test at Inyokern,43 in which some of the fuzes
were fired from a plane in ripple salvo on
HE-loaded rounds. Results are summarized
graphically in Figures 9 and 10, in which some
acceptance data are added for comparison. In
the test at Inyokern, all burst heights were
measured photographically, and they are prob-
ably more accurate than those estimated with
the camera obscura in acceptance testing.

There is no systematic evidence of sympa-
thetic functioning of HE-loaded rounds fired in
ripple salvo. In general character, the results
resemble very closely those obtained in similar
tests with the Army rocket fuzes. The main
difference is that the T-2004 burst heights are
lower (with due allowance for reflection coeffi-
cient), since the fuze was designed to have a
lower overall sensitivity in order to avoid ex-
cessive burst heights.

9 4 BOMB FUZES

9,4,1 Introduction

This section presents the results of bomb-
fuze testing, both experimental and acceptance.
Results are based in so far as possible on per-
formance of production model units; only in
cases where data from such units are inade-
quate are they supplemented by those of earlier
models.

It will be noted that certain tables in this
section are exceedingly brief. Although a much
greater amount of testing was done (which

BOMB FUZES

387

was more or less pertinent to some of the sub-
jects discussed) than is shown in the tables,
much has been eliminated in an effort to keep
the data free from extraneous factors. For

Code

□

Range
1,500 yd
2,000 yd
1,500 yd
2,000 yd

Explosive
Spotting charge
Spotting charge
High explosive
High explosive

Vertical bar indicates ± one standard error of mean.

Figure 9. Burst height as function of dive
angle, T-2004 on AR-5 for indicated mode of
firing.

example, Table 27 in Section 9.4.3, showing the
effect of vehicle on performance, includes data
for one type of fuze only, since data on other
fuzes were complicated by variations in release
altitudes or plane speeds. It is to be understood
that the pertinent data that are omitted show
satisfactory general agreement with engineer-
ing prediction but are unsuitable for analytical
purposes.

In the evaluation of performance the fol-
lowing classifications of functions are used :

Proper (Pw). A proper function is one occur-
ring because of interaction between the fuze and
the target. In acceptance work, definite arbi-
trary limits on heights were set (see appendix

to this chapter). For scoring of experimental
rounds a less rigid criterion was used. Func-
tions occurring within about twice to one-third
the mean burst height of those definitely “on
target” were scored as proper.

Low or Late (L). A low function is one
occurring below the lower limit set for proper
function.

Early (E). An early function is one occur-
ring too soon to be called proper.

Dud (D). A round in which no function
occurs is classified as a dud.

Note that all results are for release in level
flight unless specified otherwise.

Code

Dive

angle

55°-60°

40°

20°-25°

30°

Speed

(mph)

350

310-340

Range

(yd)

1,500-2,000

700-1,700

Proving

ground

Inyokern

Dahlgren

Figure 10. Cumulative burst height distribution
for various dive angles, T-2004 on AR-5.

9*4,2 Safety and Arming

General Remarks

A great deal more attention was given to
tests of the arming characteristics of bomb
fuzes than in the case of other fuzes. The ex-

388

ANALYSIS OF PERFORMANCE

perimental data on this subject are very volu-
minous. An exhaustive analysis of arming per-
formance is given in reference 22. The prin-
cipal results of this analysis are summarized in
this section. The basic data on arming per-
formance may be found in detail in reference
65.

The extensive testing on the arming mecha-
nism was primarily due to

1. The compromise between two contradic-
tory objectives. The first objective was to post-
pone arming as long as possible in order (a)
to protect the bombing plane from fragments
from bombs exploding upon arming, or (b)
to prevent the fuze from operating on other
friendly planes in a deep formation. The second
objective was to have arming occur as soon as
was reasonably safe in order to allow level or
dive bombing from low altitudes.

2. The inherent spread in arming values due
either to conditions of use or variations in man-
ufacturing tolerances.

The method of operation of the arming
mechanism has been described in previous
chapters (cf. Chapters 4 and 5). Here we are
concerned primarily with the results of tests
on the overall mechanism and of tests on the
effect of the different parameters which cause
variations in arming, i.e., (a) effective pitch
of vanes of the windmill, (b) effective airflow
around the nose of the bomb, (c) angular rota-
tion of the detonator rotor to arming, and
(d) reliability of the detonator contacts and
the rotor locking pin.

Pertinent Features of the
Arming Mechanism

1. Air Travel. The windmill-driven arming
mechanism yields air-travel-to-arming for a
given bomb and rotor-setting which is roughly
independent of altitude of release and plane
speed.

Effect of Release Altitude. To within the de-
gree of experimental accuracy necessary for
determination of rotor-settings, the air-travel-
to-arming has not been found to be affected by
the variation in air density between different
altitudes of release (3,000 to 20,000 ft). In any
event the most precise data are required only
for the low-level bomb releases, such as are

used for rotor-setting calibration tests ;
changes caused by higher altitudes are not im-
portant on account of the added safety avail-
able through use of the delayed arming device.
Further consideration of high-altitude releases
is given later in this section.

Effect of Plane Speed. Variations in plane
speed have somewhat greater effect. Wind tun-
nel calibrations of vane speed versus wind
speed performance of various windmills and
turbines indicate a decrease in air travel at
higher plane speeds. For the 10-bladed metal
vanes and 3-bladed plastic vanes, the effect is
less than the normal spread between units re-
leased under similar conditions (less than 3
per cent per 100-mph change in plane speed on
the average). For the turbine-driven type, in-
creasing the plane speed from 200 to 300 mph
appears to decrease air travel by about 10 per
cent, while a decrease to 150-mph plane speed
results in a 15 per cent increase. It should be
remembered, however, that the separation be-
tween the bomb and the launching aircraft is
less at the higher airspeeds, even with the
same air travel for the bomb. For example,
after 4,000 ft of air travel, the separation
between plane and M-30 test bomb is 2,100 ft
at 200-mph release but only 1,250 ft at 300-mph
release. The last fact is more important from
the point of view of safety than the relatively
smaller change in air-travel-to-arming.

Effect of Bomb. The air-travel-to-arming for
all fuze types is greatly modified by the aero-
dynamic characteristics of the bomb. The air-
flow about the nose of the vehicle affects the
rotational speed of the windmill or turbine to
a much greater extent than it influences the
trajectory of the bomb. For example, the tur-
bine-type fuzes (T-82) on a large bomb (M-66)
may travel more than twice the distance to
arming than on the small test bomb (M-30)
with the same rotor setting. This fact is quite
consistent with safety requirements. Data on
air travel ratios between various bombs is
given in a later part of this section, under
“Arming Performance of Typical Fuzes.”

2. Operation and Use; Rotor Setting Methods.
Setting of the rotors is accomplished by either
of two methods: (a) counting of vane revolu-
tions from the electric arming position or (b)

BOMB FUZES

389

measuring the angle of the slot in the slow-
speed shaft from the mechanical arming posi-
tion. In the first method the vanes are rotated
backward from the electric arming position by
an electric motor attached to a mechanical
counter. Electric continuity provides an indica-
tion of the electric arming position. In the sec-
ond method, the angle between the slot in the
slow-speed shaft and a reference point on the
rotor housing collar is set according to the indi-
cations of a mechanical gauge. Knowledge of
the reduction ratio of the gear train and the
angular separation between the slot in the
slow-speed shaft and the reference point on the
housing collar, when the rotor is in the electric
arming position, allows conversion of setting
specifications from one system to another. Com-
parative merits of the two methods are dis-
cussed in reference 22.

Methods of Release; Low Altitudes. Fuzes
are supplied with an arming pin which blocks
rotation of the vanes. The arming pin is held
in place (while in the bomb bay) by the free
end of an arming wire, the other end of which
is fastened to the plane. When the bomb is
dropped, the arming wire pulls out of the arm-
ing pin, releasing it and permitting the vanes
to rotate in the wind stream.

Methods of Release; High Altitudes; Use of
Arming Delay Device T-2. The arming mech-
anism of the fuze proper is not generally set to
give air-travel-to-arming greater than about
4,000 to 5,000 ft. When air travel in excess of
this is desired, a supplementary arming delay
device (T-2) is employed (see Figure 1, Chap-
ter 4). This device prevents expulsion of the
vane blocking pin until the bomb has fallen
through a predetermined distance along its tra-
jectory. The device may be adjusted manually
to yield up to 20,000 ft of additional air-travel-
to-arming for the fuze.

Settings for Given MinSAT

Statistical Method. In the production of VT
bomb fuzes, the determination of rotor settings
has been based on the requirement of a specified
minimum safe air -travel-to-arming [MinSAT]
on the smallest bomb on which the fuze is to be
used (M-30 in most cases). This MinSAT may

be defined ideally as a lower limit , below which
no fuze will ever become armed. Owing to vari-
ations within the range of currently permissible
manufacturing tolerances, the air-travel-to-
arming for an individual fuze from a given lot
may be found to differ from the mean value for
the lot by as much as several hundred feet when
dropped in the field. It is not practicable, there-
fore, to set the fuze rotors so as to yield a mean
air travel equal to the specified MinSAT, since
in that case about half of the fuzes would be-
come armed before traveling the MinSAT. Ac-
cordingly, a safety factor is introduced in the
form of a tolerance distance to be added to the
MinSAT in order to obtain the mean air travel
for which the rotors should be set.

To establish the proper value for this toler-
ance distance, it is necessary to determine the
relative frequency of occurrence of units with
air travel short of the mean by any given
amount. The functional form of this distribu-
tion of frequency may be developed mathe-
matically from the assumption that a given
deviation from the mean air travel is caused by
the random superposition of a great many
small independent deviations, each presumably
due to a variation of some physical characteris-
tic of the fuze from its average value for the
lot. The frequency distribution resulting from
these assumptions is the normal error law,
which the observations appear to follow quite
closely. This law does not, however, define an
absolute lower limit for the possible values of
air travel. In practice, therefore, the ideal defi-
nition of MinSAT given above must be modified
to denote a limit below which only a certain
negligibly small percentage of the units become
armed. With sufficiently large test samples,
the location of this limit for any given produc-
tion lot could be determined by a count of the
units of extreme short air travel. Practically,
however, such a procedure is inapplicable on
account of the magnitude of the test samples
that would be required. It is evident that in
small test samples, units with extreme charac-
teristics will seldom appear, and their scarcity
will make direct count estimates of their prob-
able frequency very unreliable. Such extreme
units, however, are to be expected in much
greater numbers in the much larger produc-

390

ANALYSIS OF PERFORMANCE

tion lots. It becomes necessary, then, to infer
their presence through an application of fre-
quency distribution theory to observations on
the more numerous, more nearly typical units
of which the small test samples are mainly
composed.

This is accomplished by determining the ap-
propriate numerical values of the dispersion
parameter for the normal frequency distribu-
tion (the standard deviation) from field test
observations of samples of the various fuze
types. The use of these values for extrapolation
from the mean air travel, according to the
theoretical distribution formula, then gives the
air travel limit which only the selected negli-
gible percentage of units will fail to exceed
before arming takes place. The MinSAT is
estimated from the mean air travel by the fol-
lowing procedure.

Consider the mean air travel as the sum of
three quantities: (1) the MinSAT, (2) a mul-
tiple of the standard deviation in air travel for
the given fuze type, and (3) a supplementary
allowance.

1. The MinSAT is understood for this pur-
pose to mean a fixed lower limit of air travel,
below which only a negligibly small percentage
of nondefective fuzes have a chance of arming.

2. The multiple of the standard deviation
represents a minimum permissible difference
between the MinSAT and the mean air travel
for the given lot corresponding to the adopted
rotor setting. The selection of the multiple
chosen is based on the condition that on the
average, only 1 per cent of the individuals in a
normally distributed population will deviate
from the mean for the entire group by more
than this multiple of their standard devia-
tion.

3. The supplementary allowance is included
principally to account for probable differences
between the mean air travel for the given lot
corresponding to the adopted rotor setting and
the mean air travel for the lot from which the
test sample was drawn which determined the
adopted rotor setting. This type of difference
may be thought to originate from sources of
variation which were not operating during the
limited range of tests used for deriving the
statistical parameters. Experience confirms

the probable existence of such factors. To eval-
uate the exact nature of such long-term pro-
duction variations (which have the effect of in-
creasing the expectation of large deviations)
would require a program of more extended
testing and wider scope than present practice
includes. Continued need for the use of a sup-
plementary safety allowance is indicated by
the large deviations from the nominal test re-
lease conditions which are likely to occur in
actual service use. A liberal margin of safety
(200 ft on the average) is generally allowed to
compensate for all of these effects.

The air travel limits between which certain
specified percentages of fuzes may be expected
to become armed are given in Table 21.

Table 21. Air travel limits (ft) .

Standard deviation of air
travel (ft)

100

150

200

250

MinSAT

3,600

Mean air travel (50%

armed)

4,030

4,145

4,260

4,375

95^% armed

4,200

4,400

4,600

4,800

Range: MinSAT to 95^%

armed

600

800

1,000

1,200

Estimates for other values of standard devia-
tion of air travel may be made by interpolation
in the table or by use of the formulas :

Range (MinSAT to 95% per cent armed) =
200 ft + 4.0 X (standard deviation of air
travel) .

MinSAT to mean air travel = 200 ft + 2.3
X (standard deviation of air travel).

Mean air travel to 95V2 per cent armed =
1.7 X (standard deviation of air travel).

For other values of MinSAT and for any
vehicle, the standard deviation of air travel
may be taken as approximately proportional to
the mean air travel.

Proving Ground Check. Proving ground tests,
conducted for purposes of production quality
control, are intended as an overall check on arm-
ing performance under conditions simulating
Service use as closely as possible. Specifications
for loading acceptance tests for VT bomb fuzes
(phase 1, see Section 9.8) provide that all of a
sample of fuzes from a given lot dropped under
certain plane speed and altitude conditions

BOMB FUZES

391

(which correspond to an air travel to ground
less than the rated MinSAT by certain tolerance
amounts) must fail to function if the lot is to
pass. Inspection criteria of this nature are in-
dispensable for effectively guarding against oc-
currence of defective units. Defective units are
defined here as those subject to essentially un-
predictable types of variation in air-travel-to-
arming such as are caused by nonrandom ele-
ments not covered in the overall method of
analysis developed earlier in this section. Ex-
amples of such exception factors are: (1) an
improper choice of rotor setting, (2) a blunder
in setting the rotor, or (3) the sudden appear-
ance of some previously unencountered type of
defect in a mechanical part. Frequency distribu-
tion sampling theory, which presupposes a con-
siderable degree of homogeneity in manufactur-
ing production, is inapplicable to the prediction
of such sources of error.

The sampling technique employed for accept-
ance testing depends simply upon the direct
enumeration of infrequently occurring types,
and, like all tests of this nature, provides little
information on the probability of occurrence
of similar rare types in other samples from
similarly composed lots. This property, as
already noted, is inherent in tests of this sort.
It is an unavoidable consequence of the limita-
tion imposed by the test design upon the num-
ber of observable specimens with the charac-
teristic about which information is sought.
Furthermore, with the usual acceptance test
procedure, the minimum permissible air travel
that may be observed for a unit (before reject-
ing the lot containing it) is not in practice
rigidly fixed. Instead, the air travel limit de-
fining a prematurely arming unit during any
particular test may lie anywhere inside the
range extending from the MinSAT distance to
a point 350 ft in advance of it. This situation
arises principally from the difficulty of main-
taining a very stringent control over proving
ground testing conditions. In addition, the ac-
tual air travel measures (based on observations
of time of flight or altitude of release) appear
to be generally subject to large random errors,
much greater than the variations which experi-
mental field tests show may be properly at-
tributed to the fuses themselves. It must be

concluded, then, that while the acceptance tests
are invaluable in screening out defective or
improperly adjusted units, the data provide
little information on the relative number of
nondefective units expected to exhibit any par-
ticular air travel to arming. In consequence,
the acceptance test rejection record is not to
be considered as a rigorous check on the ac-
curacy of the specific predictions derivable
from the MinSAT control theory presented in
Section 9.4.2. Percentages of acceptances and
rejections on the basis of air travel perform-
ance are given in summary form in Table 22.

Arming Performance of Typical Fuzes

Mean Air Travel versus Rotor Setting. The
rotor setting corresponding to a given mean air
travel for a particular fuze type is determined
by dropping a small sample of units from the
production lot. The units are wired to fire a
small explosive charge upon electric arming,
which is thus made visible from the ground.
Observation of the arming time and plane
speed, combined with a knowledge of the bal-
listic properties of the bomb, permit calculation
of the air travel to arming for each unit. From
these data, an estimate of the air travel per
vane revolution under the given conditions is
obtained, from which the rotor setting appro-
priate to a specified mean air travel may be
deduced. A description of the field testing pro-
cedures is given in Section 8.2.

The rotor calibration test procedure in prac-
tice is beset by several difficulties. First, varia-
tions in production standards between lots,
which cannot be detected by calibration tests
conducted on units all from the same lot, may
introduce large apparent errors in the choice
of rotor setting, revealed only when the units
are subjected to the acceptance testing. This
situation necessitates the addition of the 200-ft
supplemental safety allowance discussed in
Section 9.4.2 to all air travel settings. In addi-
tion, the large variation of aerodynamic pitch
(air travel per revolution) with wind speed for
certain types may, if neglected, lead to serious
errors in rotor setting estimates deduced for an
air travel different from that for which the
units were set during the calibration test. Wind
tunnel observation of windmill speed at con-

392

ANALYSIS OF PERFORMANCE

trolled wind speeds appears to be the only satis- set in order to conform with a specified

factory solution for this problem. MinSAT requires a knowledge of the standard

Typical rotor settings adopted for various deviation in air travel to be expected in Service

fuse models may be found in data sheets in use of the fuze type (Section 9.4.2). Data use-

Chapter 5. Approximate rotor settings may be ful for estimating this quantity are provided
computed from the values of the air travel per by field tests such as are used for rotor setting

Table 22. Acceptance (safety) test performance, phase 1. (A “safe” unit should fail to function.)

Mfr.

MinSAT

(ft)

Approx, air
travel to
ground*

(ft)

No. of
units
tested

No. of
functions
observed

Percentage
of functions
(failures)

Approx, tolerance*
(MinSAT minus
air travel
to ground)

(ft)

Emerson

3,600

3,500

370

+100

3,600

3,000

+600

3,100

3,300

—200

3,100

2,900

120

+ 200

2,600

2,300

+300

2,000

1,800

+ 200

Zenith

4,500

4,000

+500

3,600

330

Philco

3,600

3,500

315

+ 100

3,100

100

2,000

1,900

+100

2,000

1,900

100

+ 100

Simplex

3,600

3,500

101

+ 100

3,100

All combined ....

1,739

+ 100

* The air travel to ground (and tolerance distance) may be in error by several hundred feet.

revolution observed for various vane types in
the wind tunnel. Some results, corresponding
to an air travel of about 4,000 ft, are given in
Table 23. With bombs other than the one on
which the calibration test was conducted, allow-
ance must be made for the effect of the vehicle
upon the vane speed and consequent air travel

Table 23. Average values of air travel per vane
revolution on M-30 bomb in wind tunnel (230-250
mph wind speed).

Fuze and vane type

Air travel (ft)
per vane
revolution

6-in. Bakelite vane, bar type

1.20

6-in. aluminum vane, bar type

1.32

9-in. Bakelite vane, medium antenna

ring 1.32

55° metal, thin antenna ring

1.82

Turbine, bar type

1.47

to arming. The approximate relative air travel
for various type fuzes with the same rotor
setting on different bombs is listed in Table 24.

Spread in Air Travel. Determination of the
mean air travel for which the rotors should be

calibration. Further information may be de-
rived from analysis of wind-tunnel fuze per-
formance in combination with certain other
laboratory measurements. Both methods pro-
vide mutually consistent independent estimates
of air travel spread for the various fuze types.
The two procedures supplement each other,

Table 24. Relative air travel on various bombs
from field and wind tunnel data.

Fuze

type

Bomb type

Standard
devia-
tion of

M-30 M-81 M-64 M-65 M -66 M-56

ratios

Ring and bar
types, plastic
and metal

vanes

1.00

1.02 1.15 1.32

1.58

1.48

±0.02

Bar type, tur-
bine vane

1.00

1.02 1.24 1.48

2.32

1.87

±0.03

i.e., field testing duplicates Service use condi-
tions more directly while wind-tunnel test re-
sults are less affected by experimental error.
Comparative merits of various test methods for

BOMB FUZES

393

determining air travel spread and application
of observational data to their calculation are
discussed in reference 22, in which numerical
examples of air travel spread for various fuze
types are given. A few typical examples for late
production models (when set for 4,000-ft mean
air travel) follow.

Standard Tolerance
deviation deviation
(16% limit) (1% limit)

Manufacturer

Type

(ft)

Emerson

T-90

150

350

Zenith

T-51-E1

150

350

General Electric

T-89

175

410

Philco

T-89

230

540

Westinghouse

T-82

250

580

Examples of frequency distributions of the
deviations from mean air travel to be expected
of individual units may be found in reference
28. Comparisons with the theoretical normal
frequency law show sufficiently good agreement
between the observed and theoretical distribu-
tions to justify use of the normal law in calcu-
lating MinSAT tolerances (Section 9.4.2). An
example of such a comparison is given in Fig-
ure 11.

Figure 11. Distribution of deviations in air
travel to arming for bomb fuzes and corre-
sponding normal distribution fitted to data.

Factors Affecting Consistency of Air Travel
Performance. A brief discussion has already
been given both of the principal fuze character-
istics designed to produce constant air-travel-
to-arming under a variety of field conditions
and of the extent to which this requirement is
found to be fulfilled in practice. There remain
still to be considered the observed lack of con-
stancy of air travel performance under con-

stant field conditions and the principal factors
causing these accidental differences between
units of similar design. A comprehensive study
of this phase of air travel performance, based
largely on wind-tunnel observation, is con-
tained in reference 22.

1. Vane speed variations. Differences in air-
travel-to-arming between fuzes set to perform
identically under the same conditions are ac-
counted for principally by differences in wind-
mill or turbine speed. Variations in construc-
tion of. the detonator contacts, which cause in-
advertent differences in rotor setting under
some circumstances, are secondary in impor-
tance. The standard deviation in vane speed
between units is found to be a certain fixed per-
centage of the vane speed which is a character-
istic of the unit type (model and manufac-
turer) and is virtually unaffected by changes
of vehicle or plane speed. This relationship
leads to an approximate proportionality be-
tween spread in air travel and mean air travel.

The exact mechanical components of the gen-
erator vane and shaft system in which lack of
uniformity in manufacture is most effective in
producing the observed spreads in vane speed
have not been completely identified. Measures
of generator shaft torques and of dimensions
of certain external mechanical features of the
fuzes have thus far accounted for only a negli-
gible portion of the entire observed spread in
air travel.

The later revised production models of Emer-
son and Zenith show a great reduction in ran-
dom variations in vane speed between units.
In wind tunnel tests, these types performed
with vane speed standard deviations (and par-
tial air travel standard deviations thereby in-
duced) of about 1^/2 Per cent. Under the same
conditions, regular production models by the
same manufacturers exhibited almost twice
this spread, while variation in other manufac-
turers’ models ranged up to three or four times
as much.

2. Rotor setting errors. The next most im-
portant factor in causing spread in air travel is
a lack of uniformity in detonator contact spring
construction which leads to random setting
errors in rotor-shaft arming angles. This cir-
cumstance introduces an increase in the stand-

394

ANALYSIS OF PERFORMANCE

ard deviation of air travel over that due to
vane-speed variation ranging from about 15
to 30 ft. Errors in the actual gauging of the
shaft angles contribute only a minor portion
of this quantity.

3. Testing errors. Another important consid-
eration is the experimental field test observa-
tional error, which is responsible for a further
(apparent) increase in standard deviation of
air travel amounting to from 10 to 30 ft, the
effect being the greater for the better models
and shorter air travels. The last-mentioned
source of variation does not exist in actual op-
erational use. It may be entirely eliminated
by conducting the tests on the units in a
wind tunnel, without danger of thereby in-
troducing any new or more undesirable sources
of error.

9 4,3 Ring-Type Fuzes

Performance under Acceptance
Test Conditions

General Remarks. The analysis of perform-
ance of fuzes under acceptance test conditions,
which provides the field performance data
given in Section 5.5, is based on the metal parts
acceptance tests. These tests were carried out
on a much larger scale and under conditions
much more like combat conditions than were
the tests of the completely loaded fuzes, i.e., the
ammunition lots (see Section 9.3.1 for defini-
tions of metal parts and ammunition accept-
ance tests) . Where the metal parts tests show
significant variations in performance it would
be possible, theoretically, to make some allow-
ance for these variations in predicting the per-
formance of the ammunition lots. However, the
composition of ammunition lots is highly vari-
able. The lot size ranges from a few hundred
to several thousand fuzes, and a lot may con-
tain from one to nearly a dozen metal parts
lots, or fractions thereof, widely spread with
respect to time of production. Furthermore, the
loading of the fuzes is known to have intro-
duced factors in one or two cases that were
absent in the metal parts testing. Therefore,
although the properties of the ammunition in
its final form are of primary practical interest,

it is not feasible to attempt a quantitative pre-
diction of the effect of variations observed in
metal parts testing. For this reason, and be-
cause most of the statistically significant vari-
ations are not of much practical importance,
the data have not been analyzed as exhaustively
as would be possible.

In calculating the scores from acceptance
test data, the scores obtained with rejected lots
are included, since what is required is an esti-
mate of the quality of the product. As with
most sampling procedures, rejections were
almost entirely the result of random sampling
fluctuations, and elimination of the reject
scores would yield an overestimation of the
quality of the fuzes.

For the sake of simplicity, the Army Ord-
nance designations for the complete fuzes are
used in discussing the performance of the metal
parts assemblies, although strictly speaking
the Signal Corps system of AN/CPQ desig-
nations should be used (see Section 5.5).

Conditions for Acceptance. The procedure
for conducting metal parts acceptance tests is
summarized as appendix material to this chap-
ter in Section 9.8. The following test conditions
hold except where deviations are noted :

Nominal altitude of release: 10,000 ft.

Nominal true airspeed at release: 200 mph.

Test vehicle

For Brown carrier fuzes:

M-81 (260-lb) fragmentation bomb or
M-88 (220-lb) fragmentation bomb.

For White carrier fuzes:

M-64 (500-lb) general purpose [GP]
bomb.

A double sampling formula was used with
the proper functioning requirements stipulated
so that a negligible fraction of the metal parts
lots would be rejected as long as the manufac-
turing quality was above an 80 per cent proper-
functioning level.244

Summary of Data for Various Groups of
Lots. In Table 25 there is given the average
score and function height for each of several
groups of metal parts lots of each of the fuzes.
The lot group number (first column) is used
for reference purposes in the discussion, and
to identify the metal parts lots to which the
data apply, with the aid of auxiliary Table
26.

SECREf'

BOMB FUZES

395

For the most part, the reason for the group- Where it is not apparent in the table, the

ing of the lots is apparent, namely, differences reason for the grouping is given in the discus-
in plane speed, or target factor. The target sion of the performance of the fuzes,
factor (second column) is the reflection coeffi- Effect of Test Conditions on Performance.

cient of the terrain expressed as a percentage. There is no reason to anticipate any perceptible
It is to be understood that the target factor is difference in scores on account of a 40-mph dif-

Table 25. Metal parts acceptance test results.

Mean

Standard

Lot

Number

burst

error

group

Target

units

Per cent

score

height

mean

No.

factor

tested

(ft)

Brown carrier fuzes

T-50-E1 and

T-89 (Philco)

542

0.7

Ice

226

1.0

2.0

T-50-E1 (Philco-Simplex)

387

0.9

Ice

2.2

T-91 (Philco)

3.6

204

0.4

680

0.6

1.9

T-91

(GE)

499

0.9

1.7

12*

259

1.2

13*

106

1.9

M-168 (T-91-E1) (Emerson)

14*

289

1.2

15*

170

1.3

White carrier fuzes

T-50-E4 (Emerson)

370

0.9

680

1.0

279

1.1

Ice

3.0

736

0.8

21*

4.7

T-92 (Emerson)

1,000

0.7

23*

4.0

T-92-E1 (Emerson)

3.1

25*

2.3

4.8

* 240 mph nominal true airspeed at release.

approximately 81 (water surface at Aberdeen
Proving Ground) unless otherwise stated
(other numerical values apply to Jefferson
Proving Ground). The reflection coefficient of
ice depends upon a number of conditions which
were not recorded in detail for these tests.

ference in plane speed, and an inspection of
Table 25 shows no consistent indication of such
a difference.

With the rather wide limits of burst height
that were used in the classification of proper
functions, no perceptible effect of target factor

396

ANALYSIS OF PERFORMANCE

on score would be anticipated, and none is ob-
served in the table.

These conditions (plane speed and target
factor) are therefore disregarded in estimat-
ing the overall score for each fuze.

Table 26. Identification of lot group numbers
listed in Table 25. (Duplication in lot number in-
dicates part of the lot was tested under conditions
of the particular group number.)

Lot

group

No.

Composed of metal parts lots

CPR 43 through 59, 62 through 79, 102, 104,
106

CPR 80 through 101

CPR 108, 110, 112, 114, 116, 118, 140

CPR IS through 23S, 25S, 29S, 30S

CPR 24S, 26 S, 27S, 28S

CPR 103, 105, 111

CPR 166 through 177

CPR 107, 109, 113, 115, 117, 119, 120 through
139, 141 through 155, 157 through 165

CPR 178 through 180

CG 1 through 23

CG 24, 25, 27, 31

CG 27 through 39, 41, 44, 45

CG 40, 42, 43, 45 through 48

CEX 3001 through 3014, 3019, 3021, 3022

CEX 3015, 3018, 3020, 3023 through 3027

CEX 1 through 19, 21

CEX 20, 22 through 50

CEX 51 through 74

CEX 75 through 80

CEX 81 through 84, 86 through 104, 106
through 109, 112, 113, 114, 116, 117, 118,
120, 121, 122, 124, 125, 126, and all even
numbered lots through 144, 148

CEX 146

CEX 105, 110, 111, 115, 119, 123, 127, and
all odd numbered lots through 195, 201

CEX 205, 207

CEX 178, 184

CEX 182, 184, 176, 178, 180

CEX 174

In considering the mean burst heights, the
situation is different. Inspection of the pre-
dicted isoburst height charts (Section 5.5) in-
dicates that increasing the plane speed from
200 to 240 mph should increase the burst height
by something like 10 ft. The theory of opera-
tion of the fuzes indicates that a reduction in
reflection coefficient from 0.81 to 0.65 should
reduce the burst height by 10 to 20 per cent
or by something like 5 ft for these data. Such
differences are large in comparison with many
of the tabulated standard errors of mean burst

heights, and one might therefore hope to ob-
tain a good check on the predictions from these
data. Unfortunately, there are present in the
data certain variations in burst height that
cannot be accounted for quantitatively. Atten-
tion is called to specific cases presently. At this
point it is sufficient to remark that these vari-
ations are of a magnitude similar to the pre-
dicted differences just mentioned. Therefore
no precise check on the theory from these data
is possible.

In view of the presence of uncontrolled fac-
tors in the data, it is very desirable to estimate
the burst height performance from all of the
data rather than from selected groups. If the
effect of reflection coefficient (other than that
of ice) is tested by determining burst-height
ratios between lot groups for which all other
conditions are presumably equal, and these
ratios are weighted in accordance with the
amount of data and averaged, it appears that
on the whole the effect of target factor (other
than ice) is nil. This same procedure indicates
that the average target factor of ice, as it
affected these tests, is about 92, a result which
agrees with an unpublished investigation of
this matter. The general conclusion is that some
uncontrolled factors are present in the data
that give burst heights at Jefferson Proving
Ground that appear to be higher than would be
expected from the Aberdeen data. This factor
might be a difference in the systematic com-
ponents of the errors in burst-height measure-
ment at the two locations. In order to be con-
servative in combining data to obtain overall
average burst heights for each fuze, the effect
of target factor (other than ice) is neglected.

A like analysis shows that the overall aver-
age effect of the 40-mph increase in plane speed
is to add about 7 ft to the mean burst heights.
This value is in reasonable agreement with
engineering prediction and it is used in re-
ducing all lot group mean burst heights to a
200-mph basis before calculating average per-
formance for each fuze type, although to be
strictly correct a slightly different correction
should be made for each fuze type.

Performance of T -50-El and T-89. The first
six lots of Philco production were tuned on a
load equivalent to the M-64 bomb and gave

BOMB FUZES

397

fairly satisfactory performance on this bomb,
although an excessive number of duds was
observed. The main cause for the duds had been
found in the type-approval test to lie in faulty
detonator contact springs. When these lots
were tested on the M-81 bomb, a very large
number of early functions was observed and
production was halted temporarily. Specifica-
tions were changed to require laboratory test-
ing on a load equivalent to the M-30 bomb, and
the amplifier was changed from the No. 8 to the
No. 10 type, which became standard for the
T-50-E1.

Acceptance testing was resumed with the
M-81 as a standard vehicle, but the dud per-
formance remained poor through lot 42. Omit-
ting retest scores (which would throw undue
weight on exceptionally poor lots), the overall
performance on M-81 for this series was

Per cent Lots

491

8-42

102

24, 27, 33, 35, 37, 42

(reworked lots)

The score for the six lots that failed the re-
test and were subsequently reworked showed
a significant but not entirely satisfactory re-
duction in the occurrence of duds. The lots that
were not reworked were, for the most part,
loaded into ammunition lots through PA —
180 — 15; reworked lots were loaded later.

The performance of the balance of the T-50-
E1 and T-89 production is given in Table 25,
items 1 through 5. There are no markedly sig-
nificant differences in these scores, which give
an average of

Proper 83 per cent

Late 0 per cent

Early 13 per cent

Dud 4 per cent

Number tested, 1,278
Mean burst height, 35 ft

able uniformity and give the following pooled
estimates.

Per cent

Philco

Both

Proper

Late

Early

Dud

Number

963

917

1,880

In early GE production some difficulty with
detonator contact springs caused a relatively
large number of duds, but the situation was not
serious enough to warrant special discussion as
in the case of early T-50-E1 production. The
Philco T-91 fuzes appear to have benefited from
the special attention that had to be given to
this problem in the earlier model. No T-91
metal parts production lots were rejected.

There is a rather large difference (15 ft) be-
tween the mean burst heights of lot groups 6
and 7, which cannot be explained by any dif-
ference between the characteristics of the lots
as measured in the laboratory. The only re-
corded difference in test conditions that might
be pertinent is in the test vehicle ; group 6 was
tested on the M-81 and group 7 on the M-88
bomb. However, there is no known difference
between the properties of these bombs that is
large enough to account for so large a differ-
ence in burst heights. The observed difference
is probably due to a chance combination of fac-
tors, no one of which is alone sufficient to
account for the difference in heights. Group 7
appears to be more inconsistent than group 6
with the other groups, 8 and 9, of Philco T-91,
but there is no reason to reject it from the
overall estimate of burst heights:

Product
T-91 (Philco)
T-91 (GE)
T-91 (both)

Mean burst
height (ft)
37
44
40

The burst height of the Simplex product is
slightly higher than that of Philco but the
difference is of little practical importance. The
overall average burst height over Aberdeen
water is 35 ft.

Performance of T-91. The scores of the vari-
ous lot groups of both Philco and General Elec-
tric Company [GE] production show reason-

The higher burst height of the GE fuzes may
be due in part to greater r-f sensitivity (18 as
compared with 16 v for Philco), but the un-
accountably low burst height of Philco group
7 makes a fair comparison impossible.

Performance of M-168 (T-91-E1) . This was
a late model fuze, and no more need be said
here about its performance other than that it

398

ANALYSIS OF PERFORMANCE

leaves little to be desired. Overall performance
(lot groups 14 and 15) is as follows.

Proper
Late
Early-
Dud

Number tested, 459
Mean burst height, 50 ft

92 per cent

0 per cent
7 per cent

1 per cent

Performance of T-50-EU and T-90. The
scores obtained with the various lot groups
(16 through 21) in this series are fairly uni-
form. The lower dud scores appearing in later
production are probably due mainly to improve-
ments in detonator contact springs. The excess
in early functioning of group 20 as compared
with groups 16 and 17 may be due in part to
factory rejected and reworked assemblies from
T-92 production that were absorbed into T-90
production. Excess earlies in group 19 are not
so significant because this is a relatively small
group. The variations in early and dud scores
happen to be of a compensating nature, so that
the proper function scores are quite uniform.

Proper

Late

Early

Dud

Number tested, 2,153

77 per cent
1 per cent
19 per cent
3 per cent

Of the approximately 130 metal parts lots of
these fuzes, only 3 per cent were rejected.

As in the case of Philco T-91, there appears
with these fuzes a difference between mean
burst heights of lot groups that cannot be ac-
counted for quantitatively. The difference is
about 10 ft (see groups 16, 17, and 18). There
was an upward trend in carrier frequency, to
the extent of about 1 me through part of the
production represented by groups 16 and 17,
but both theory and correlation tests using
field data indicate that this could account for
no more than 2.5 ft. As in the case of T-91, the
difference is probably due to a combination of
factors, and no rejections are desirable in esti-
mating the overall average burst height, which
is 39 ft.

Performance of T-92 and T -9 2-El. The T-92
is the only production fuze that exhibited gen-
erally unsatisfactory scoring performance. The
initial performance did not appear to be too
bad, and it is suspected that a certain change

in the fins of the test bombs (see “Effect of Re-
lease Conditions” in this section) may have had
something to do with the deterioration in per-
formance. This does not appear to provide a
full explanation ; a more basic explanation con-
cerns an unusual dependence (at White fre-
quency on the M-64) on the electric resistance
between the bomb and its fin (see Section
2.13.2). The broader pass band of the T-92
amplifier made the effect more critical than in
the T-50-E4. Quality of parts and workman-
ship were apparently not at fault. Overall per-
formance was as follows:

Proper 57 per cent

Late 1 per cent

Early 34 per cent

Dud 8 per cent

Number tested, 1,034
Mean burst height, 33 ft

Of the 45 metal parts lots produced of the T-92,
over 60 per cent were rejected.

Only a few lots of the T-92-E1 were produced
before the end of hostilities. Overall perform-
ance was as follows :

Proper 79 per cent

Late 0 per cent

Early 18 per cent

Dud 3 per cent

Number tested, 102
Mean burst height, 40 ft

Burst Height Distribution Characteristics.
For a given type of fuze the spread in burst
heights increases approximately in proportion
to the mean burst height when the latter is in-
creased by a change in such factors as reflection
coefficient, altitude of release or plane speed.
The evidence on this subject is summarized in
“Burst Heights under Other Conditions” later
in this section.

The general character of the distribution of
bursts obtained under supposedly constant
testing conditions is shown for three fuzes in
Figures 12, 13, and 14. The lot group numbers
of Table 25 are given to identify the sources
of data. The largest lot groups are used, with-
out regard to target factor. It will be noted
that the distributions are much the same in
general character for all three fuzes. For some
purposes the cumulative distribution curves
(Figures 15, 16, and 17) are more useful. For
practical purposes, these distributions may be

SECRET

BOMB FUZES

399

assumed to be linear when the cumulative per-
centage is plotted on a probability scale and
the burst height on a logarithmic scale. On

Figure 12. Distribution of burst heights of
Philco T-50-E1 fuzes over water (lot group 1).

account of the approximately constant propor-
tionality between the spread and the mean
burst height, this type of distribution curve is

Figure 13. Distribution of burst heights of
Philco T-91 fuzes over land (lot group 8).

simply displaced without change in slope when
conditions alter the mean burst height. For a
detailed study of distribution characteristics,
see reference 35.

Summary. The overall estimates of perform-
ance derived above are summarized in Table
27 for those fuzes that gave reasonably uni-

Figure 14. Distribution of burst heights of
Emerson T-50-E4 fuzes over water (lot group
20).

form and satisfactory performance. Excluded
from the table are the first 42 metal parts lots
of T-50-E1 that gave a high incidence of duds,
and the whole of the T-92 production which

Figure 15. Cumulative distribution of burst
heights of Philco T-50-E1 fuzes over water (lot
group 1).

was rather uniformly unsatisfactory on ac-
count of early functioning. Almost no T-92
metal parts were loaded into ammunition lots.

400

ANALYSIS OF PERFORMANCE

In comparing the performance of Brown and
of White carrier fuzes it is apparent that the
proper functioning performance of the Brown
class is distinctly better than that of the White
class, mainly on account of a lower incidence of

Scores under Other Conditions

Uniformity of Performance on Various Ve-
hicles. Although the effect of vehicle size upon
performance may be quite marked for fuzes of
the ring type (if detuning occurs), observed

Table 27. Summary of metal parts acceptance test performance. (Burst heights are for a water target of
reflection coefficient 0.81 and a plane speed at release of 200 mph.)

Fuze

Make

No.

tested

Per cent scores

L E

Mean burst
height (ft)

Brown carrier fuzes

T-50-E1

Philco

and T-89

and Simplex

1,278

T-91

Philco

963

917

both

1,880

T-91-E1

Emerson

459

White carrier fuzes

T-50-E4

and T-90

Emerson

2,153

T-92-E1

Emerson

102

early functions. This difference would be much
more marked if the T-92 were included. On the
other hand the overall incidence of duds, 4 per

Figure 16. Cumulative distribution of burst
heights of Philco T-91 fuzes over land (lot group
8).

cent, is slightly higher for the Brown carrier
fuzes, and would be still higher if the first 42
lots of T-50-E1 were included in the comparison.

performance of the T-91 shows very good con-
sistency for several different bomb sizes, as
shown in Table 28. There is an apparent im-

Figure 17. Cumulative distribution of burst
heights of Emerson T-50-E4 fuzes over water
(lot group 20).

provement in performance on the M-64 bomb
but the size of the sample is too small to give
the trend real statistical significance.

BOMB FUZES

401

Table 28. T-91 fuzes, performance versus vehicle

(from 10,000 ft at 200 mph).*

Vehicle

Per cent

L E

Total
No. of
units

M-30

M-57

M-88, -81

798

M-64

* Reference 66, acceptance tests.

Effect of Release Conditions. (1) Altitude.
In Table 29 are listed scores as a function of
altitude of release, without regard to plane
speed. A previous examination of the data had
shown the latter to have no appreciable effect
upon performance. It will be noted that results
are fairly uniform (with one exception) with a
tendency toward slightly poorer scores for high-
altitude releases. The exceptionally high early-
function scores for the T-92 units from 10,000
and 20,000 ft should be considered in the light
of the discussion in preceding section under
“Performance of T-92 and T-92-E1” (pertain-
ing to the data in Table 25) .

2. Train spacing. The dependence of per-
formance in train release upon the spacing of
bombs is influenced by size of the bomb and
the use of the delayed arming device. In Table
30 are listed separately results without delays
and those with delays. As may be seen, there
is no indication of serious effect of train release
upon dud score, and the effect upon early func-
tioning decreases as the train spacing increases.

The effect of spacing along with that of the
arming delay may best be seen in Figure 18,
where the per cent of early functions which oc-
curred sympatheticallyf is plotted against inter-
val between bombs at release.

Because the data are meager, the random
variations in results mask to some extent the
real effects of the delays and of bomb size as
shown in results of the bar-type fuze. There the

f Early functions which were judged to have been
caused by malfunctioning of neighboring units were
called “sympathetic” functions. For the purposes of
scoring, where photographic data were not available,
functions within 0.5 sec or less of the original function
were scored as sympathetic. Photographic evidence
showed that for the most part, these functions occurred
nearly simultaneously.

evidence is strong that arming delays cut down
appreciably the sympathetic functioning and
that as the spacing increases, the sympathetic
functions decrease rapidly.

In Table 30, “int’l” stands for “intentional
early.” In many train drops, one or more units
were set to function on arming, at times later
than normal arming time. This insured the oc-
currence of one or more early functions in order
to test the possible response of the other
armed fuzes in the train to it. In the table,
“sym” means sympathetic function; (Sym X
100) /E is the percentage of earlies functioning
sympathetically as plotted in the accompanying
figure.

Oddments. (1) Washers: Hand versus wrench
tightening 18 During early testing of bomb fuzes,
the general practice became established of
mounting units to bombs using a %2-in- lock
washer and tightening with a wrench. When it
became apparent that the elimination of the
wrench would be desirable, a new-type spring
washer was introduced and the units were
tightened by hand only. This method of mount-
ing proved to be satisfactory for use with both
the ring-type and the bar-type fuzes.

A representative score of 83 per cent proper
for T-91 fuzes released under standard condi-
tions was obtained using this method.

2. Navy base plates.18 In Naval aircraft
launching operations, a fuze protective device
was used, consisting of two parts: (1) a metal
sleeve, which was released when the bomb was
dropped, and (2) a cruciform plate 7 in. in di-
ameter, which served to hold the sleeve in posi-
tion. This plate remained between the bomb
nose and lock washer throughout the entire
flight. Tests were made to see what effect the
plate had on fuze performance. The results
follow :

No. of

Unit

Vehicle

Alt.

Speed Pw L

Total %Pw

T-50-E4*

M-64

10K

200

6 0

T-50-E4

M-64

10K

200

9 0

T-91

M-64

10K

200

18 0

18 100

* There is no known explanation for the difference in performance
in the two tests of T-50-E4 units.

3. Army guide plates.18 The performance of
T-91 units, assembled on bombs with Air Corps
arming-wire guide plates, was tested. No sig-

402

ANALYSIS OF PERFORMANCE

nificant effect was found. The results are as 5. Fin insulators ,75 In an attempt to reduce

follows: the effect of fin upon the incidence of early

^ functions, tests were made with pressboard

Unit Mfr. Vehicle Alt. Speed Pw L E D Total Pw insulating spacers inserted between the outer

T-91 Philco M-88 10K 200 ll 0 1 0 12 92 rim of the fin and the bomb. Somewhat contra-

T-91 Phllco M“64 10K 200 12 0 0 0 12 100 dictory results were obtained. In Table 32 are

4. Fin thickness .74 Sometime during the listed results for three types of units dropped

spring of 1945, excessive early-function scores under comparable conditions on M-64 bombs,

Table 29. Effect of release conditions (single release).

Altitude

(ft)

Speed

(mph)

Per cent

L E

Total Biblio-

No. of graphical

units reference

9K-10K*

160-200

Unit T-91
71

Vehicle M-57

24 67

255

20K

234-240

Unit T-91

Vehicle M-81 & M-88

37 15, 68 acceptance tests

12K

200

10K

200

798

200

9K-10K

150-200

Unit T-91

Vehicle M-6U

98 69

255

20K-21K

218-240

Unit T-50-E4
81

Vehicle M-81 & M-88

48 70

12K

200

10K

200

100

20K-22.5K

210-240

Unit T-50-EU
64

Vehicle M-6U

15K

200

12 71 acceptance tests

10K

200

1,131

7.5K

200

20K

240

Unit T-92

Vehicle M-6U

20 15, 72 acceptance tests

10K

200

1,246

200

2.5K

200

* K represents 1,000-ft units of altitude.

were obtained in tests of White-frequency units
mounted on M-64 bombs. It was found that the
metal of the fins on these bombs was thinner
than the usual 0.081-in. material. Tests con-
ducted with M-64 bombs having fins of various
thickness gave the results in Table 81.

The above results with the 0.081-in. fin are
in agreement with acceptance results of the
same unit, i.e., 57 per cent proper for 1,000
units. Acceptance tests were made on the M-64
having nominal fin thickness of 0.081 in.

both with and without the insulators. The dif-
ference in scores for the T-92 units (disregard-
ing duds) could occur fortuitously one or two
times in a hundred. However, the scores for the
T-92-E1 and T-50-E4 units show no statistically
significant differences.

6. Delayed arming device .7(5 The delayed arm-
ing mechanisms were designed as safety de-
vices and as a means of improving performance
of fuzes dropped from high altitudes. The de-
vices have been tested primarily to obtain air-

BOMB FUZES

403

travel data; a few were tested for performance, (from inexactitudes of both electric character-

The scores of the latter tests follow. istics and conditions of release).36 In addition,

% % mean observed burst heights from actual field

Unit Vehicle Alt. Speed Pw L E D N Pw E testing15’ 78 have been spotted in with brackets

T-50-E4 M-64 20K 240 14 0 4 0 18 78 22 indiratine- the standard error of the mean (+1

T-50-E1 M-81 20K 235 19 0 3 0 22 86 14 bleating the standard error ol the mean < ±l

T-91 M-81 20K 234 5 0 0 0 5 100 0 standard deviation) and the number of rounds

Table 30. Performance in train,73 HE-loaded bombs, ring-type fuzes (from 10,000 ft at 200 mph.)

No. of

Early

No. of

Sym X 100

Unit

Vehicle

trains

units Proper Low

Int’l

Other

Dud

Sym

Without DAD *

Interval: 15 ft

T-50-E10

M-81

29f 19 0

T-50-E1

M-64

12$ 8 0

Interval: 50 ft

T-50-E10

M-81

48$ 33 0

T-50-E4

M-64

18 6 0

Interval: 100 ft

T-50-E10

M-81

24 18 0

With DAD

Interval: 15 ft

T-91

M-81

35§ 13 0

T-50-E1

M-64

12 10 0

Interval: 50 ft

T-91

M-81

36f 24 0

T-91

M-64

12 9 0

T-91

M-65

6 10

Interval: 100 ft

T-91

M-64

18 14 0

T-91

M-65

6 10

Interval: 150 ft

T-91

M-65

4 0 0

* DAD = delayed arming device,
t 2 unaccounted for.
t 1 unaccounted for.

§ 5 unaccounted for.

Although these results indicate no appreci-
able effect of the delay upon performance, it
should be noted that train drops of T-51 units
made at Eglin Field indicated marked reduction
of earlies by the use of the delay.

Burst Heights under Other Conditions

Effect of Altitude. Predictions of burst
heights for various ring-type fuzes as a func-
tion of release altitude for level flight at 200
mph have been made using the method described
in reference 25 and laboratory data given in
Chapter 5. In Figures 19, 20, 21, and 22, pre-
dicted heights are represented by solid lines;
the dotted lines represent an appraisal of the
cumulative error involved in the calculations

upon which the means are based. Figure 23 is a
plot of mean observed heights (photographic)

Table 31. Effect of fins on performance.*

Unit

Pw L

E D Total

%Pw

Fin-metal thickness: 0.073 in.

T-92

33 0

34 5 72

T-92-RGD

8 0

10 0 18

T-50-E4

8 0

9 1 18

Fin-metal thickness: 0.081

in.

T-92

42 0

23 6 71

Fin-metal thickness: 0.105

in.

T-92-RGD

14 0

2 2 18

T-50-E4

24 0

6 0 30

* All rounds were dropped from 10,000 ft at 200 mph. The
difference between the performances with the 0.073- and 0.105-in.
fins is highly significant statistically.

SECRF/lf

404

ANALYSIS OF PERFORMANCE

versus predicted heights for these tests plus a
number of earlier tests in which various experi-
mental model fuzes were involved.77 Each mean
was obtained from a single test.

Table 32. Effect of fin insulators on performance.

Unit

Insulator

Total

T-92

Used

T-92

Not used

570

100

1,000

T-92-E1

Used

T-92-E1

Not used

T-50-E4

Used

T-50-E4

Not used

738

163

959

It may be seen that despite rather large dis-
crepancies in some cases between observed and
predicted heights, for practical purposes the
agreement is satisfactory.

Effect of Vehicle. The design of the ring- type
fuze is such that height of burst is affected by
the bomb with which the fuze is used. In Table
33 are given mean observed heights h for vari-
ous fuze-missile combinations dropped under
standard conditions (from 10,000 ft at 200
mph) along with an estimated standard error
of the mean S-h. For comparison are listed also
heights for the given combinations predicted by
the method outlined in reference 25 using lab-
oratory data given in Chapter 5. Agreement be-
tween the two values of heights is satisfactory
if allowance is made for a possible 15 per cent
error in prediction, as estimated on the basis of
reasonable discrepancies in laboratory data and
conditions of release.30

Table 33. Effect of vehicles on function height,
ring-type fuzes (from 10,000 ft at 200 mph).

Unit

Mfr.

Predicted
height h

Vehicle (ft) (ft)

(ft)

T-50-E1

Philco

M-81, M-88

0.5

T-50-E1

M-81, M-88

3.2

T-50-E4

Emerson

M-64

0.7

T-91

Philco

M-30

2.3

T-91

Philco

M-81, M-88

0.9

T-91

Philco

M-64

1.1

T-91

M-57

4.9

T-91

M-81, M-88

0.9

T-92

Emerson

M-64

0.6

References 15 and 79, acceptance tests.

Effect of Train Release. Visual and photo-
graphic observations indicate that when fuzes
are dropped in train, a certain number of func-
tions within the proper range are due, at least
in part, to the functioning of neighboring fuzes.
This is evidenced by an occasional stacking up
of bursts with successive fuzes functioning at
increasing heights. The effect of this sympa-

TRAIN INTERVAL (FT)

1_J I I I I I I I I 1 ■! I 1 -1

0 10 20 30

40 50 60 70 80 90 100 IIO 120 130 140 150 160 17

TRAIN INTERVAL (FT)

Code

Fuze

T-50-E10

T-50-E4

□

E-50-E1

■

T-50-E1 with arming delay

▲

T-91 with arming delay

Figure 18.

Effect of train spacing and of arm-

ing delay upon sympathetic functioning of ring-
type bomb fuzes on HE loaded M-81, M-64, and
M-65 bombs.

thetic functioning among bursts within the
proper range would be expected, on the whole,
to raise the mean height of function.

From the data tabulated in Table 34, there
appears to be no dependence of burst height on
this effect. It appears either that sympathetic
functioning occurs less often than suspected

ECRET

BOMB FUZES

405

from visual observation8 or that tolerances al-
lowed in the manufacture of fuzes permits
variation in burst height sufficient to partially
mask the effect. This is further borne out by a
Navy test of T-91 fuzes, released in train in a
dive of 30 to 45 degrees (at 250 mph) , in which
with intervals at release of 90 ft the rounds

height of T-89. Vertical bar covers ±1 standard
error of mean.

scored as sympathetic propers in one train
occurred higher than the regular propers and
in another train lower.

Table 34. Effect of train interval on function height,80
HE trains, ring- type fuzes.*

Fuze: T-

50-E4

T-50-E10

T-91

’-91

Train

Bomb : M-64

M-81

M-64

interval

(ft)

(ft) (ft)

(ft)

±6.4

±3.0

±3.5

±3.1

±3.4

±3.3

100

±4.3

±3.0

±0.07

±0.87

±1.1

* The term h = mean burst height. The term = estimated standard
deviation. Heights for single release drops are based on data for inert
fuzes.

Spread in Burst Height as a Function of
Mean Burst Height. In Figure 24 are plotted
values of standard deviation of burst height as
a function of the mean burst height. Each point
(except a few which cover acceptance testing
of particular fuzes) represents the results of

£ Studies have been made of photographic data to
determine actual spacing between bursts which from
one location appear to be sympathetic functions. It was
found that in many instances these spacings were of
such magnitude as to preclude the possibility of inter-
action.

an individual test of ring-type fuzes (involving
on the average about 10 units). The line shown
was determined by the method of least squares.

Although inspection shows that some points
fall far from the line, a general trend is defi-

Figure 20. Effect of release altitude on burst
height of T-90. Vertical bar covers ±1 standard
error of mean.

nitely indicated. For the purposes of rough
estimates of burst height spread, a measure of
this trend has been found useful. As determined
from the slope of the straight line, the standard
deviation is about 0.4 times the mean height.34

Figure 21. Effect of release altitude on burst
height of T-91. Vertical bar covers ±1 standard
error of mean.

9 4 4 Bar-Type Fuzes

Performance under Acceptance
Test Conditions

General Remarks and Lot Group Data. To the
statements made in the first three paragraphs
of Section 9.4.3 concerning the acceptance test-

406

ANALYSIS OF PERFORMANCE

in g of and presentation of data on ring-type
fuzes nothing need be added for bar-type fuzes
except the information that the test vehicle was
the M-81 (260-lb) fragmentation bomb. The lot
group data are given in Table 35 and the group
composition in Table 36.

height of T-92. Vertical bar covers ±1 standard
error of mean.

Effect of Test Conditions on Performance.
The theory of operation of bar-type fuzes pre-
dicts that under the conditions of these tests the

garded in estimating overall burst-height per-
formance.

There is no reason to expect the scores to be
affected by either plane speed or target factor
within the range of the test conditions. It could
be shown from an analysis of lot-to-lot per-

Figure 23. Observed versus predicted burst
heights, ring-type bomb fuzes. Each point repre-
sents one test. Line is the least-squares straight
line of best fit.

formance within groups 1 and 7 that there was
a general downward trend in early functioning

Table 35. Metal parts acceptance test results.

Lot

group

No.

Target

factor

Number

units

tested

Per cent score
L E

Mean

burst

height

(ft)

Standard

error

mean

(ft)

T -51-El and T-51-E2 (Zenith)

590

113

1.0

1,605

0.5

170

1.4

677

0.9

M-166 (Zenith)

112

2.2

3.2

M-166 (Emerson)

300

1.6

149

1.6

T-712 (Zenith)

100

1.9

* Indicates 240-mph nominal true airspeed at release.

effect of changes in plane speed on burst height
would be of the same order as the tabulated
standard errors of the means. Nothing in the
data is to the contrary and this factor is disre-

during the early production periods involved.
The relatively high early-function scores of
these groups thus happen to be associated with
•higher target factors than the next lot group

BOMB FUZES

407

in each case. Target factors, plane speeds, and
the trends just mentioned are all disregarded
in calculating overall performance.

In contrast to the case with the ring-type
fuzes, the mean burst heights of the bar-type
lot groups are closely associated with the values
of target factor, but the relation does not ap-

Code Fuze

O T -50-El, T-50-E4

X T-91, T-91-E1

A T-92

Figure 24. Standard deviation versus mean
burst height, ring-type bomb fuzes. Each point
represents one test.

pear to be one of strict proportionality. The
burst height does not appear to change quite as
rapidly as the target factor. A small and some-

Figure 25. Distribution of burst heights of
Zenith T-51-E1 fuzes over land (lot group 2).

what arbitrary allowance for this situation is
made in adjusting the group burst heights to
a common target factor of 81.

Burst Height Distribution Characteristics.
The comments made in Section 9.4.3 on “Burst

Height Distribution Characteristics” apply
equally well to bar-type fuzes. Evidence on the
relation between spread and mean burst height
of bar-type fuzes is given in the following para-
graph. The ratio of spread to mean is in gen-
eral smaller for bar-type than for ring-type

Table 36. Group identification for Table 35.
(Duplication in lot number indicates part of the lot
was tested under conditions of the particular group
number.)

Group

number Metal parts lots

1 CHU 1 through 50, 70

2 CHU 51 through 119, 121 through 158

3 CHU 194 through 198, 205, 206, 227, 228, 230

4 CHU 158 through 191, 199 through 204

5 CHU 5002, 5003, 5005 through 5009

6 CHU 5001, 5004

7 CEX 5001 through 5010, 5014 through 5018

8 CEX 5011 through 5013, 5019 through 5023

9 CHU 192, 193

fuzes, and the distribution curve (see Figure
25) is more symmetrical. A cumulative distri-
bution curve is given in Figure 26.

Figure 26. Cumulative distribution of burst
heights of Zenith T-51-E1 fuzes over land (lot
group 2).

Summary of Performance. Apart from the
variations mentioned in the preceding para-
graph “Effect of Test Conditions on Perform-

408

ANALYSIS OF PERFORMANCE

ance,” the performance of the lot groups of
each manufacturer is very uniform. The overall
results are given in Table 37. The proper func-
tioning performance of the Zenith product is
higher than that of Emerson, mainly on ac-
count of the difference in early functioning. It
is reasonable to assume that if the Emerson
production had continued for a longer period,
lower early function scores would have been
obtained.

No metal parts lots of bar-type fuzes were
rejected.

Table 37. Summary of metal parts acceptance
test performance. (Burst heights are for a water
target of reflection coefficient 0.81.)

Mean

burst

Number Per cent scores height
Fuze Make tested P L E D (ft)

T-51-E1,

M-166

Zenith

3,195

1 9 <1

110

M-166

Emerson

449

2 15 <1

110

Scores under Other Conditions

Uniformity of Performance on Various
Vehicles. The design of the bar-type fuze is
such that performance should be relatively in-
dependent of the size of vehicle used. The field
test results on inert-loaded vehicles, listed in
Table 38, show no statistically significant dif-
ferences in scores. There is a suggestion of
somewhat impaired performance for the T-82
on M-57. However, in the absence of any
known physical basis for poorer performance
on this bomb, it appears that the lower score
should be attributed merely to sampling fluctu-
ations. In any event, the effect is not serious.
Numerous tests were made also of T-51 on
other miscellaneous missiles (fire bombs, chem-
ical bombs, etc.) and uniformly good perform-
ance was obtained. Table 39, giving scores from
the Eglin Field service test of T-51 units on
HE-loaded vehicles,49 indicates further the con-
sistency of performance from vehicle to vehicle
as well as that between inert- and HE-loaded
rounds.

Effect of Release Conditions. In Tables 40
and 41 are listed results of single drops as a
function of release altitude for inert- and for

He-loaded vehicles. In two cases only do the
proper scores fall below the 80 per cent mark ;
both of these are for releases from 30,000 ft.

Table 38.
loaded.

Performance versus

vehicle,

inert-

Vehicle

Per cent

Piv L E

Total

No.

T-82 fuze, released from 10,000 ft at 200 mph*

M-30

M-57

M-88 )

M-81 S

480

M-64

M-65

M-66

100

M-56

100

T-51 fuze,

released from

10,000 ft at 200 mph*

M-57

522

M-88 (

M-81 S

877

M-64

M-56f

100

M-56$

100

* Reference 81, acceptance tests.

t These tests of T-51 on M-56 were made from various release
altitudes (6,000 to 10,000 ft) and over various targets at Aberdeen
and Eglin Field.

t This group represents T-51 fuzes with reduced sensitivity (about
|One-half normal) prepared specifically for use on M-56. The average
burst height was 74 ft over the water target at Aberdeen. The
reduced sensitivity fuze later carried the designation T-712.

Table 39. Performance versus vehicle, HE-loaded.

Vehicle

Per cent

L E

Total

No.

T-51 fuze, released from

10,000 ft at 225 mph

M-30

100

M-81

M-64

T-51 fuze, released from 10,000 ft at 175 mph

M-30

M-81

M-64

The remaining scores are consistently high and
indicate little dependence of performance upon
altitude of release.

Performance in Train Release, with Special
Reference to the Effect of the Arming Delay
Device. The effect of train release upon the
performance of bar-type fuzes is similar to that
found with the ring-type fuze. Dud scores are
affected but little, while early functioning is in-
creased as train intervals decrease.

Variations in score with bomb size and with

BOMB FUZES

409

use of arming delays may be seen in Table 42.
The graph in Figure 27 shows how the use of
delays and increased spacing cut down sympa-
thetic early functioning.

Table 40. Effect of release conditions (single re-
lease), bar-type fuzes, inert bombs.

Altitude

(ft)

Speed

(mph)

Per cent
L E

Total

No.

T-82 Fuzes82

Vehicle:

M-88 and M-81

22.5K*-25K

235-250

10K

200

480

200

T-51 Fuze88 1

Vehicle:

M-88 and M-81

24K-25K

225-250

20K

210-250

10K

200

877

255

T-51 Fuze 84 Vehicle :

M-6U

10K

200

255

100

* K denotes altitude in thousands of feet,
t Includes acceptance test results.

Table 41. Effect of release conditions (single re-
lease), bar-type fuzes, HE-loaded bombs.49

Altitude

(ft)

Speed

(mph)

Per cent
L E

Total

No.

T-51 -El Fuze-

* Vehicle

: M-

-30

30Kf

225-260

10K

175-225

150-200

T-51 -El Fuze

Vehicle.

: M->

30K

230-270

10K

175-225

150-200

T-51 -El Fuze

Vehicle .

: M- 1

30K

230-270

10K

175-225

150-200

* Since a previous examination of the data had shown that the
effect of the T-2-E1 device upon performance (in single release) was
not appreciable, results with the delay were included in the tables
above.

t K denotes altitude in thousands of feet.

The effectiveness of the arming delays in im-
proving performance in train releases is out-
standing. For example, in the worst possible
cases, M-64 in salvo and minimum train, the
use of the delays elevated the proper function
score from 57 per cent to a level (86 per cent)
that is almost identical with the performance
of the fuzes in the metal parts acceptance tests

(single release). The results indicate, there-
fore, that the arming delays are highly effec-
tive in eliminating sympathetic early function-
ing. Indeed there is strong evidence that the
proper use of arming delays obviates the need
for certain limitations on train spacing that
were considered to be of considerable impor-
tance until these tests were performed.

Oddments.1* (1) Washers, Hand versus
Wrench Tightening. As with the ring-type fuze,
the first established method of mounting bar-
type fuzes to bombs was with a %2-in. lock
washer and tightening with a wrench. The

TRAIN INTERVAL (FT)

Without

With

Bomb

arming delay

M-30

M-81

▲

M-64

□

Figure 27. Effect of train spacing and of arm-
ing delay upon sympathetic functioning of T-51
fuze on HE loaded M-30, M-81, and M-64 bombs.

effect upon performance of the use of a lock
washer and tightening by hand only was tested.
The following typical results show no indication
of deterioration in performance when these

410

ANALYSIS OF PERFORMANCE

fuzes are mounted hand-tight instead of wrench- No deterioration in performance had been

tight. anticipated, and none was observed.

gpeed 3. Delayed Arming Device. Except for train

Alt. (mph) Vehicle Pw L E D N %Pw drops of T-51 units in the Eglin Field service

T-51 Fuze test, only a few bar-type fuzes equipped with

20K 240-250 M-88, -81 22 0 l 1 24 92 delayed arming mechanisms were tested for

10K 200 M-88 30 0 0 0 30 100 normal approach function. The results are listed

T -82 Fuze below.

10K 200 M-88 21 0 2 0 23 91

Unit Vehicle Alt. Speed Pw L E D N fyPw %E
2. Army Guide Plates. The performance of T_82 m-81 10K 200 9 0 l 0 10 90 10

T-51 units assembled on bombs with Air Corps T-51 M-81 10K 200 8 l l 0 10 80 10

arming-wire guide plates, was tested. Twelve T"51 M"81 25K 22°"250 H 1 0 0 12 92 0

rounds on M-88 from 10,000 ft at 200 mph gave These results are too meager to indicate

100 per cent proper function. much improvement in performance. However,

Table 42. Performance in train,49 HE-loaded bombs,* T-51-E1 fuzes (from 10,000 ft at 200 mph).

No. of

Sym X 100

Vehicle

trains

fuzes

P L

Sym

With no ay'ming delay

device

Salvo

M-30

23 1

M-81

3 0

M-64

22 0

Minimum train

M-30

38 0

M-81

38 2

M-64

27 0

50- ft interval

M-30

37 0

M-81

40 2

M-64

22 1

100- ft interval

M-30

44 0

M-81

37 0

M-64

22 0

With T-2-E1 arming delay device

Setting

: 7 Div. on M-30, -81.

6 Div.

on M-61t

Salvo

M-30

47 0

M-81

42 0

M-64

54 0

Minimum train

M-30

39 0

M-81

40 1

M-64

48 0

50-ft interval

M-30

45 0

M-81

39 0

M-64

42 1

100- ft interval

M-30

39 0

M-81

43 0

M-64

20 1

* Table combines trains dropped over water and land,
t Low-order detonation.
t One function on impact.

BOMB FUZES

411

the drops in train at Eglin Field showed a
marked reduction in early functioning by the
use of the delay.

Burst Heights under Other Conditions

Effect of Altitude. Dependence of burst
height upon altitude of release is shown in Fig-
ure 28. Curves of heights predicted by the
method given in reference 33 are shown. Mean
burst heights from field testing have been
spotted in along with an estimate of their

dicated differences in fuze sensitivity, and
when differences in release conditions obtained.

Effect of Train Release. The possibility of
sympathetic functioning for the ring-type fuzes
in the region of proper burst heights has been
discussed in Section 9.4.3. If sympathetic func-
tioning did occur in the proper function zone,
one would expect to find the burst height in-
creasing with smaller train spacing. However,
the mean burst heights of bombs fuzed with the
T-51, as in the case of the ring-type fuzes, do

Figure 28. Effect of release altitude on burst height of T-51 fuzes on M-30, M-81, and M-64 bombs.

standard error and the number of units in-
volved. It will be seen that if allowance is made
for a discrepancy of about 10 per cent in pre-
diction, the agreement between predicted and
observed heights is reasonably good.

Effect of Vehicle. The effect of vehicle on
height of burst of bar-type fuzes is shown in
the data tabulated below. Table 43 gives burst
heights for single releases over water from
10,000 ft at 200 mph. Table 44 gives the ratio
of the burst heights for the fuzes on various
bombs to those on the M-81. (Data are based on
experimental field tests at Aberdeen, service
tests at Eglin Field, and metal parts acceptance
tests.) In deriving these ratios adjustments
were made when laboratory measurements in-

not appear to be affected by train spacing. The
mean burst heights for several different bombs
and for difference train releases are presented
in Table 45.

Spread in Burst Height as a Function of
Mean Burst Height. A plot of spread in burst
height (see Figure 29) as a function of mean
burst height for the bar-type fuze shows much
the same trend as that for the ring-type fuze
(see Section 9.4.3). In Figure 29 the standard
deviation from the mean is used as the measure
of spread, each point representing the results
of one test. Although the scatter from the line,
which was determined by the method of least
squares, is in some cases quite marked, for
practical purposes a measure of the trend has

412 ANALYSIS OF PERFORMANCE

Table 43. ]

Effect of vehicle on function height (bar- type fuzes).

Fuze

M-30

h* Si f

M-57
h Si

t-h 00

00 00

'-ss

M-64
h Si

M-65
h Si

M-66
h Si

M-56
h Si

T-51

T-82

142J =*= 14f
107 ±4.6

121 ±1.4
111 ±5.9

111 ±0.9

120 ±1.4

75 ±3.3

99 ±4.5

77 J ±8J

57 ±3.5

44 1 ±3t

57 ±4.2

1574 =*=12t

67 ±7.7

* h = mean burst height in feet.

t Sfi = estimated standard error of the mean. Reference 85, acceptance tests.

| These heights estimated from Table 44; no field data available for standard release conditions.

Table 44. Ratio of burst heights of various bombs
to burst height of M-81 with T-51 and T-82 fuzes.86

Bomb

T-51

T-82

M-30

1.28 ± 0.10

0.94 ± 0.05

M-57

1.00 ± 0.05

0.94 ± 0.06

M-88

1.00 ± 0.05

1.00 ± 0.03

M-17

0.60 ± 0.07

M-64

0.73 ± 0.03

0.86 ± 0.05

M-65, -79

0.69 ± 0.10

0.50 ± 0.03

M-66

0.40 ± 0.06

0.53 ± 0.04

M-56

1.37 ± 0.08

0.49 ± 0.05

was superimposed on the other to form a box
1 ft deep. The bottoms consisted of %-in. ply-
wood panels 6 ft long and 2 ft wide. Trenches
were excavated 6 in. beneath the surface, the
boxes inserted and leveled, and a sand parapet
built around the upper 6-in. frame. The
trenches were arranged in columns and rows,
47 each, spaced 15 ft between centers. Odd-
numbered columns had the long dimension of
the trench in an east-west direction, while

Table 45. Effect of train interval on burst height,49 HE-loaded bombs fuzed T-51-E1.

Type of

Minimum

50-ft

100-ft

100-ft train

Bomb

release

Salvo

train

double susp.

h* Si f

h Si

(ft)

M-30

128

±7

121

±5

127

±4

142

±5

152

±8

M-81

130

±3

108

±3

126

±4

145

±6

M-64

±5

±4

±3

* The term h = mean burst height (released over water, from 10,000 ft).
fThe term Si = estimated standard error of the mean.

been found useful. As determined from the
slope of the line, the standard deviation is
about 0.23 times the mean height.34

Effectiveness of Air Burst

Enhanced Fragmentation Effect

Against Moderately Shielded Personnel. Test-
ing was done at Eglin Field to determine the
relative effectiveness of air-burst and contact-
burst bombs against moderately shielded per-
sonnel (both T-50 and T-51 fuzes were used).
Bombs were dropped over an effect field
700x700 ft. The field contained 2,209 replica
trenches, constructed as follows. The sides con-
sisted of two rectangular frames 6 ft long and
2 ft wide constructed of lx6-in. pine. One frame

those in even-numbered columns had the long
dimension in the north-south direction.

In scoring the results, a casualty was defined
as one or more “large” perforations through
the wooden lining of a trench. In the classifi-
cation of hits as “large” or “small,” a probe
approximately %6X%6 in* in cross-sectional
area was used. Relatively few perforations
were as small as this probe. Figures 30 and 31
show results for single releases from 6,000 ft
at 165 mph (chosen to give maximum accuracy
of aim and at the same time give the same
striking angle as a 10,000-ft, 200-mph release)
of 19 M-81 bombs VT-fuzed, as follows: 14
M-64 bombs VT-fuzed; 6 M-81 bombs contact-
fuzed (instantaneous) ; 5 M-64 bombs contact-
fuzed (instantaneous) for 3 degrees of shield-
ing. Ten M-1A1 clusters of M-41 bombs, con-
tact-fuzed (instantaneous), were included in

BOMB FUZES

413

the test. Hits for the various degrees of shield-
ing were included in the count, as follows.

Shielding Hits Counted

0 in. On sides and bottom

6 in. On lower 6 in. of sides and bottom

12 in. On bottom only

No significantly differing results were ob-
tained in additional releases from 12,000 and
20,000 ft.

In Section 9.2.3, attention was called to the

Complete analysis and interpretation of the
results is somewhat lengthy; reference may be
made to the Army Air Forces Board report,45
where the following conclusion is drawn.

“Under the conditions of this test and for
equivalent airplane loads of properly function-
ing bombs, air-burst M-81 or M-64 bombs are
about ten times as effective in producing frag-
ment casualties as are the same bombs or the
20-lb M-41 fragmentation when contact-burst.’'

Code Fuze

A T-82, T-82-E1

X T-51, T-51-E1

Figure 29. Standard deviation versus mean burst height, bar-type bomb fuzes. Each point represents
one test.

significance of “zero shielding” in the assess-
ment of results obtained on this effect field.
From the foregoing description of the field, it is
clear that, except for a bomb that strikes in-
side a trench, a fragment cannot register a hit
unless it is traveling in a downward direction
from a level above that of a parapet tlrat sur-
rounds a trench. Many of the fragments from
a contact-fuzed bomb fail to satisfy this condi-
tion. The results of the test are therefore not
applicable to a case of troops lying exposed on
a mathematically flat plane. The zero-shielding
condition approximates more closely a case
where troops are lying on a terrain with fre-
quent irregularities of an average height or
depth of about 6 in.

Against Unshielded and Shielded Personnel
and Against Unshielded Materiel. The British
carried out an evaluation of air-burst bombs
against a composite close support target.87
American fuzes (T-50 type) and bombs
(M-64) were used. Division 4 cooperated in the
tests, operating through the London branch of
OSRD.

The target consisted of unshielded trenches
and simulated prone and entrenched personnel.
About 200 trenches 2x6 ft and 1 ft deep were
randomly located in an area about 500x1,000 ft.
Centered in the bottom of each trench was a
16x46x1/2-im target board, simulating the vul-
nerable area of either three men standing or
crouching in a deep trench or one man lying in

414

ANALYSIS OF PERFORMANCE

a shallow trench. Two boards (10x60xV2 in.)
nailed together as an inverted trough were
placed on the ground near each trench, simulat-
ing prone soldiers on reasonably level ground.
Thirty-six trenches were scattered at random
in the center of the target area. One or more
perforations through a board target counted

Figure 30. Casualties as function of burst
height of M-81 fragmentation bomb for several
degrees of shielding. Vertical bar indicates ±1
standard error of mean.

as a casualty. In the case where the boards in
trenches represented three men in a deep
trench, the number of perforations were
counted, and on a probability basis, scored as
one, two, or three casualties. Boards destroyed
by blast within 50 ft of the burst were scored
as complete casualties. Trucks were carefully
examined after each bomb burst and only those
damaged to the extent that rear-echelon repair
was necessary were counted as casualties.

Results were expressed in terms of lethal
areas. Comparison tests with contact-burst
bombs were not made at the same time but pre-
vious tests had established vulnerable areas for
them under reasonably similar conditions. A
summary of the results and comparison data
are shown in Table 46.

Against Unshielded Materiel and Deeply En-
trenched Personnel. An extensive test to deter-
mine the effectiveness of air-burst and contact-
burst fragmentation and incendiary bombs
against typical enemy defense fortifications
was conducted at Eglin Field.47 Because of the
complexities involved in the analysis of the re-
sults, details are herein omitted.

From the analysis, it appeared that for

totally unshielded trucks and light materiel, a
plane load of air-burst bombs was about 80
per cent as effective as one of contact-burst
bombs. However, the advantage may be more
apparent than real in view of the following
considerations :

1. When the results of the test were assessed,
it was not known that double suspension of
VT-fuzed M-30 and M-81 bombs was prac-
ticable. Twenty-four M-30’s and 22 M-81’s were
used as full VT-fuzed loads in B-17 bombers;
it now appears that 34 bombs could have been
carried without sacrificing good performance.

2. The assessed effectiveness might be
greater if account were taken of the likelihood
that aircraft would be dispersed in revetments.

Figure 31. Casualties as function of burst
height of M-64 GP bomb for several degrees of
shielding. Vertical bar indicates ±1 standard
error of mean.

As for deeply entrenched personnel (depth
of shielding was 5 ft) , so slight was the dam-
age done by either air or contact burst that no
assessment of relative effectiveness could be
made.

Miscellaneous Eglin Field
Testing of Effectiveness

During the later stages of World War II,
much interest developed in fire bombing with
napalm-gasoline gel. Spectacular results were

BOMB FUZES

415

obtained by dropping fuel tanks of this ma-
terial from fighter planes flying at such a low
altitude that the gel sloshed over a large area
immediately after the tanks were ruptured by
impact. Under conditions where low-level at-
tack was too dangerous, high-altitude releases
of contact-fuzed vehicles gave poor results be-
cause a large fraction of the gel remained in
the crater. A number of tests were performed
with VT-fuzed vehicles to overcome this diffi-
culty. In all cases, the performance of the VT

the possible increased lethality of large blast
bombs when air burst. Early in World War II
the British prepared a number of special fuzes
to provide air bursts on their 4,000-lb high-
capacity [HC] bomb. Burst heights were set
for around 200 ft, which was believed to be the
optimum height of function. These bombs were
dropped over enemy territory and the damage
assessed by photographic coverage. The results
showed a decrease in area of demolition and a
small increase in area of minor blast damage92

Table 46. Advantage ratios in favor of 500-lb bombs fuzed T-50.

Men

Height

Men in

prone

of burst

deep

shallow

without

Mechanical

Target

(ft)

trenches

cover

transport

Effectiveness relative to surface

4.0

3.7

1.3

1.0

bursts

3.7

5.3

1.2

0.4

Effectiveness relative to a

1.8

1.2

1.4

1.6

method yielding 50% air

1.7

1.3

0.7

bursts using 0.6-sec train

spacing*

Lethal or vulnerable areas (sq

5,200

5,600

25,000

38,000

ft)

4,800

8,000

24,000

16,000

* This method consisted of dropping

a stick of four bombs fuzed with

No. 44 pistol (a pressure-activated fuze).

Usually two bombs of

the train would function on impact and

the other two

would be air burst.

actuated by the blast of

the others. Tests with this arrangement

had been performed previously.

The British results were in substantial agreement with the Eglin Field results described when due allowances are made for the differences
in scoring.88 The lower ratios of effectiveness reported by the British for air burst were due primarily to the allowances made for blast. This
allowance increased appreciably the lethal area of the surface-burst bombs.

Because of the minor differences in effectiveness against personnel for 10- and 36-ft burst heights and the appreciably greater effective-
ness of the former against mechanical transport, the British prefer the lower burst heights. In their official requests for American fuzes for
their operational use, they specified burst heights of the order of 10 ft.

fuzes was satisfactory, and the main problem
was to find a container that was available in
large quantity in the theaters of operation that
could readily be modified into a bomb having
suitable bursting and ignition characteristics,
suitable ballistic characteristics, and that could
be carried economically by fighters or bombers.
One improvisation was a modified chemical
warfare M-10 (35-gal) spray tank, which gave
rather satisfactory results when fuzed T-50-
E4, or fuzed T-51-E1 in combination with a
“slow” burster.46 Considerable success was
achieved in an extensive program for the de-
velopment of a vehicle specifically designed for
fire bombing, but this program was incomplete
when terminated at the close of hostilities.

Enhanced Blast Effect

A number of experiments were performed
both in America and in England to determine

and, accordingly, interest in the air burst of
large blast bombs diminished. Following these
tests, work on the T-40 and T-43 fuze projects
(see Chapter 1 and Section 3.5) was appre-
ciably curtailed.

However, Division 2, NDRC, and British ex-
plosive experts were convinced that demolition
from blast could be increased by air burst and
set about to determine the optimum height.
Extensive small-scale89 and model-village tests90
showed conclusively that blast damage could be
increased by air burst at the proper height.
Optimum heights for a 4,000-lb bomb were esti-
mated at between 40 and 70 ft.91 Increases in
area of demolition were estimated at from 50 to
100 per cent. The conclusions were further cor-
roborated by analysis of the areas of damage
produced by a few V-l bombs which were acci-
dentally air burst in the London area.92

No full-scale tests were carried out prior to

416

ANALYSIS OF PERFORMANCE

the end of World War II to verify the above
conclusions. However, as is shown in Table 38
above, the T-51 fuze could be modified to oper-
ate reliably on the M-56 (4,000-lb) bomb. The
British also established that the T-51 fuze
could be modified to operate reliably on their
4,000-lb HC bomb.93

Enhanced Spread of Gas

A number of tests were carried out to deter-
mine the effectiveness of air-burst bombs in en-
hancing the spread of mustard-type gas. These
tests were made by the British in England and
in Anglo-American tests in Panama in simu-
lated jungle warfare. The T-51 and T-82 fuzes
were used on British 500-lb light-case [LC]
Mark II bombs. In this bomb, which contains
two bursting elements, the proximity fuze was
located in the nose and an impulse fuze was
located in the tail. The actuation of the prox-
imity fuze triggered the tail fuze so that both
bursters were effective in dispersing the gas.
The results of the tests in England have not
been published but have been communicated
verbally to Division 4.94 The results showed
that for a 50-ft average air-burst height, the
area contaminated to the extent of 1 mg per
sq m was approximately 4% times the area
contaminated by the surface burst. For burst
heights in the range 100 to 200 ft, the area con-
taminated (again 1 mg per sq m) was approxi-
mately seven times the corresponding area for
surface-burst bombs. Publication of the British
results was withheld pending an investigation
of the inflammability of mustard gas when air
burst. Subsequent tests at Panama95 showed
that mustard did not ignite when air burst.

In the tests at Panama96’ 97 in which the con-
ditions of jungle warfare were simulated, it
was desired to have the air burst just below the
level of the treetops in order to contaminate the
area under the canopy. It was found that full-
sensitivity T-51 fuzes gave air bursts in the
treetops or just above, but half-sensitivity T-51
fuzes gave burst heights just below the treetop
level and produced an optimum effect. Results
from the Panama test were :

1. Thirty bombs impacted on the target area
produced an average contamination density of
about 2 bombs (containing 350 lb of agent) per

artillery square, or about 54 tons of agent per
square mile.

2. Vapor dosages greater than 200 mg-min
per cu m were attained over about three-quar-
ters of the target area within the first four
hours after bombing. Dosages exceeding 1,000
mg-min per cu m were obtained over about one-
third of the area in this period.

3. The half-sensitivity T-51 fuze was con-
sidered a suitable and desirable fuze for the
British bomb, aircraft LC 500-lb Mark II,
charged with blister gas for use on jungle ter-
rain.

95 FUZES FOR MORTAR SHELLS

951 General

Since no VT fuzes for mortar shells reached
the mass production stage, it is not possible to
estimate the quality of performance that could
have been attained with production fuzes. Ex-
perience with all other VT fuzes showed that
performance of mass production models was
superior to that of experimental and pilot pro-
duction models. There is no reason to suspect
that experience with mortar-shell fuzes would
have been otherwise. It is, therefore, believed
that an average measure of the observed per-
formance of the more recent developmental
models may be fairly presented as a lower limit
for the quality of performance that could be
expected of production fuzes based on the de-
velopment prior to V-J Day.

The presentation of a lower limit for quality
of performance is not very satisfactory in
evaluating the potential usefulness of a devel-
opment program, and the data are therefore
presented very briefly. Specific test references
are omitted. Coverage is defined by a statement
of the characteristics of the tests performed
during a certain period that are included in the
analysis. The source material can be identified
by reference to summaries of field test re-
sults.19’ 20> 30’ 39

• •

95 2 Reliability and Burst Heights
Globe-Union T-132

Performance under Standard Conditions. A

FUZES FOR MORTAR SHELLS

417

fairly large number of pilot production Globe-
Union T-132’s were fired in what were called
“lot quality tests/’ These were fired under
somewhat similar conditions at Blossom Point
and at Clinton Proving Ground. In the sum-
mary in Table 47, the analysis is limited to
those rounds fired on M-43C shells.11

As one would expect, the fuzes differed from
lot to lot as test results indicated that changes
were necessary. Variations among the units
considered in this summary include the follow-
ing:

1. Amplifier plates: thin or thick horizontal
plates.

2. Turbine speed: high (most units) to low
(half of high speed).

3. Nose shape: flat (most units) and
rounded.

4. Thrust washers: vary in number from
one to nine.

5. Regulation circuit: series or parallel;
light or heavy loading.

grouping was necessary to consolidate the
Blossom Point data with the results from Clin-
ton Proving Ground where no distinction be-
tween early and middle functions was made.
A small fraction, about 4.3 per cent of the
proper functions given above, was classed in
earlier reports as impact functions. Because
the average burst height was in the neighbor-
hood of 10 ft, and because the principle of op-
eration of the fuze makes it improbable for the
fuze to function normally at more than three
levels between 15 and 0 ft, some functions at
water level were to be expected. Detonations
initiated at the surface may yield bursts below
the surface, on account of delay in the detona-
tor and spotting charge. It is doubtful that any
of these functions were actually caused by im-
pact, through either mechanical or electric
action.

The poorer scores for firing with charge 4
(Table 47) were being corrected by V-J Day
by the use of slower turbine speeds, better

Table 47. Summary of performance of Globe-Union T-132 for the months of June and July 1945 at Blossom
Point and Clinton.

Quadrant

Charge elevation Per cent

(M-56) (degrees) N P E D H (ft) N h*

270

184

75, 80

221

153

196

60, 65

249

300

360

150

75, 80

Overall scores

1,690

697

* Number of rounds upon which height is based.

6. Arming: settings were changed occasion-
ally but not enough to affect performance
markedly. The above variations did not cause a
statistically significant difference among the
function scores, and should not affect the func-
tion height. The absence of significant difference
may be partly due to the small number tested
with some variations, but there is no serious
objection to pooling the data for the purpose of
this summary.

In Table 47, the early functions include those
called middle functions at Blossom Point. This

h The M-43C is a combination of the M-43 body and
the M-56 tail.

thrust washers, and vertical-plate amplifiers
(see Chapter 4). These changes would reduce
early function and dud scores by reducing
breakage of the plates at high acceleration,
erratic behavior of the mechanical system at
high acceleration and speed, and explosion of
rotors by centrifugal force.

Performance after Packaging Tests. The
only special test which is pertinent to the pres-
ent discussion is that of performance after
packaging tests at Picatinny. Twelve Globe-
Union T-132 fuzes from lot GUS-17 were tested
in the laboratory before and after being sub-
jected to packaging tests at Picatinny. (The

418

ANALYSIS OF PERFORMANCE

laboratory tests showed changes in electric
characteristics which, in view of changes in
control units from the same lot not subjected to
the packaging tests, were considered caused by
aging only.) The field test of these units gave
the following results :

Quadrant

elevation

Burst

Charge (de-

Num-

Per cent height

(M-56) grees)

ber

FED (ft)

4 45

58 8 33 9

This score is not significantly different from
that in the summary above.

National Bureau of Standards T-171 Fuze

The National Bureau of Standards [NBS]
T-171 fuze was used as an experimental unit,
and many variations in electric and mechanical
systems and in power supply were used. The
summary in Table 48 includes only those units
with RC arming. The other variations found
among these units would be expected to affect
function heights obtained with M-56 Ext.1 shell
so that no reliable figure may be given for
overall performance in that respect. The aver-
age function heights for various tests on this
shell varied from 36 to 61 ft. The function
scores and heights with the M-43C shell were
not affected enough to prohibit pooling the re-
sults. Firing conditions were limited almost
exclusively to those listed in Table 48.

There are no pertinent tests made with NBS
T-171 which are not included in Table 48.

Table 48. Performance of NBS T-171 from June 1

to September 20, 1945.

Quadrant

elevation

Charge

Vehicle M-56

(de-

grees)

Per cent
P E

(ft)

M-43C 2

M-43C 4

139

M-56 Ext. 1

Overall score

241

WURLITZER T-171 AND ZENITH T-172 FUZES

Not enough rounds were fired with either of
these fuzes to obtain any idea of their probable
future performance.

1 The M-56 shell with a 2-in. rearward displacement
of the tail assembly, designed to stabilize flight of the
VT-fuzed shell.

Safety and Arming
Globe-Union T-132

General. No data are available on times or
distances to complete arming of T-132 units.
Summary data on mechanical arming are pre-
sented below. The most reliable data are prob-
ably the arming times, obtained either from
fuzes modified to function on mechanical arm-
ing [FOMA] or from fuzes so modified that the
carrier signal was extinguished momentarily
on carrier indication of mechanical arming
fCIMA] (see Chapter 8).

It should be noted that the arming times are
approximately inversely proportional to the ve-
locity during burning, so that the mechanical
arming distance is nearly independent of the
propellent charge. It follows from this that the
round-to-round variations in velocity that oc-
cur with a fixed charge are reflected in round-
to-round variations of arming time. For this
reason, it is to be expected that the propor-
tional variation in arming time would exceed
the proportional variation in arming distance.
A detailed analysis of arming performance is
not included here, since a major change in the
arming mechanism was under development
(see Chapter 4), and would undoubtedly have
been used if the fuze had gone into produc-
tion.

Arming Data. (1) Arming times. Arming
time of a FOMA unit was obtained by averaging
values obtained by several field observers with
stopwatches or by averaging the stopwatch
times to the end of the phonograph recording of
carrier modulation obtained by playing the
record several times. Arming time of a CIMA
unit was obtained by measurement of a photo-
graphic record of carrier modulation or by
averaging stopwatch times obtained by playing
the phonograph recording of carrier modulation
several times. These techniques are described
fully in the preceding chapter. Results on arm-
ing times are shown in Table 49.

2. Arming distances. An attempt was made
to measure the slant distance from firing point
to function of 47 FOMA units. The work was
complicated by photographic troubles and the
results, shown in Table 50, may be in error by
±50 ft.

FUZES FOR MORTAR SHELLS

419

NBS T-171 Fuze

In order to obtain a unit which would have
the electric and mechanical systems isolated as
much as possible, most NBS T-171 units were
made without an out-of-line element in the

Table 49. Times to mechanical arming of GU
T-132.

Charge

Arming

indi-

cation

No. of Arming time (sec)

units Max Min Mean SD

A rming

setting: 2,600 turbine

turns

FOMA

16 4.2

3.3

3.7

0.37

CIMA

25 4.3

3.5

3.8

0.36

FOMA

6 2.7

2.4

2.6

0.10

CIMA

6 2.7

1.9

2.1

0.30

FOMA

6 2.3

2.0

2.1

0.10

CIMA

6 2.7

1.9

2.1

0.30

FOMA

13 2.2

1.7

1.9

0.23

CIMA

42 2.2

1.6

1.9

0.20

Arming setting: 2,1+00 turbine

turns

FOMA

24 4.2

3.4

3.8

FOMA

23 3.8

1.6

2.0

Table 50. Slant distance to arming of GU T-132.

No. of Slant distance to arming (ft)

Charge

units Max

Min

Mean

Arming setting: 2,1+00 turbine

turns

1,290

1,070

1,170

1,200

990

1,110

powder train.

This did away with the

gear

train running through the entire length of the
unit. Arming was accomplished by an RC delay
in the firing circuit which prevented firing
until a certain time after the generator started
to provide plate voltage. This was strictly an
experimental design, not intended for Service
use. The only available datum on arming of this
unit is the time to the earliest function ob-
served in any test.

Nominal Minimum
time time

No. of R C to arming to function

units (megohm) (/-if) (sec) (sec)

231 5.0 0.6 2.3 3.9

95,4 Ranges

Since the weight of a VT mortar fuze is rela-
tively large compared with the weight of the

shell, slight changes in fuze shape and size may
result in noticeable effects on mortar shell bal-
listics. In Table 51, the effect of fuze shape on
range is indicated. It should be noted that the
effect for the M-56 shell with 2-in. tail exten-
sion is not as marked as in the case of the
M-43C ; this is due to the fact that the drag of
the M-56 extension shell is quite large, and the
weight of the shell is greater.

The weights of the fuzes, with the exception
of the PD fuze, were approximately the same.
The T-132A is a streamlined T-132. The T-132B
is slightly more streamlined than the T-132A
(see Chapter 4). Range data given for the
T-171 are based on field tests of fuzes manu-
factured at NBS. These fuzes had flat noses
similar to the T-132. The T-171 fuzes manufac-
tured by Wurlitzer had rounded caps similar to
the T-132A, and their ranges were greater than
those of NBS T-171. However, no data are
available for these fuzes at an elevation of 45
degrees. The T-172 has a loop antenna and can-
not be compared in shape to the other fuzes.

The effect of various mortar shell types,
fuzed with T-132, on ranges is shown in Table
52. Since the M-43C is lighter and smaller than
the M-56, its range is longer. The M-56 with
the 2-in. tail extension had the shortest range.
(The long tail structure, with its increased
space about the powder, caused slower burning.
In addition, the increase in projectile length
decreased the distance through which the pres-
sure from powder-burning could act. The tail
extension, however, increased the stability of
the shell.)

It should be remembered that the incre-
ments of charge used with the M-43 shell are
smaller and of a different type from those used
with the other shells. Charge 6 for M-43 is
roughly the same as charge 4 for M-56.

No adjustments were made to allow for the
effect of wind direction and velocity upon
range. Data in Tables 51 and 52 are from field
tests conducted at Blossom Point Proving
Ground only.

In Tables 51 and 52, the weights listed are
those for fuzed projectiles. These weights are
only approximate. In the field tests from which
the range data were obtained, the proper func-
tioning of the fuze was of primary interest.

420

ANALYSIS OF PERFORMANCE

The shells were cavitated for permanganate
puffs, and no special effort was made to equal-
ize the weights.

The figures in parentheses, following range
values, are the number of rounds upon which
the ranges are based.

Since in many instances the fuzes were ini-
tiated to combat so late in World War II, only
preliminary or trial usages were made. Follow-
up orders for fuzes after first trials were not
fulfilled in time to be of value.

The information presented in this section

Table 51. Ranges of mortar shells: effect of fuze.14- 30- 39

Shell: M ~Ij.SC Charge: 4

T-132

T-132 A

T-132B

T-171 T-172

PD M-52B1

Elevation

(7 lb, 10 oz)

(7 lb, 10 oz) (7 lb, 12 oz)

(6 lb, 13 oz)

45°

6485' (356)

7295' (6) 7300' (6)

Shell: M-56 + 2-in. ext.

6445' (106) *

9495'f (12)

Charge: 1

T-132

T-132A

T-132B

T-171

Elevation

(11 lb, 10 oz)

45°

2075' (5)

2170' (6)

2040' (6)

2025' (47)

* Tests

conducted at the Clinton Proving Ground indicate that ranges for

T-172 on M-43C shells, fired with charge 4 at 45° quadrant

elevation, are approximately equal to ranges for T-132 fired under the same conditions.

t Fired

at 46° elevation.

Table 52.

Ranges of mortar shells: effect of shell type (fuzed T-132).30 Elevation: 45°

M-43

M-43C

M-56

M-56 ext.

Charge

(7 lb, 12 oz)

(7 lb, 10 oz)

(11 lb, 8 oz)

(11 lb, 10 oz)

2610' (3)

2980' (39)
4315' (50)

2500' (3)

2075' (5)

3780' (3)

3910' (5)

4270' (2)

5205' (2)

5320' (40)

5095' (7)

6485' (356)

96 OPERATIONAL USES OF BOMB AND
ROCKET VT FUZES

961 General

The VT fuzes for bombs and rocketsj were
employed in combat against both the Germans
and the Japanese by both the Army and Navy
with varying degrees of success. The VT bomb
fuzes were used in general-purpose, fragmenta-
tion, and incendiary bombs, against antiair-
craft (flak) positions, air fields, trains, and
light fortifications to give maximum blast and
fragmentation effect, and to disperse incen-
diary material over buildings and troop con-
centration areas. The VT rocket fuzes were
employed in ground-to-ground, air-to-ground,
and air-to-air roles to destroy aircraft hangars,
aircraft on the ground, aircraft in flight, and
to disperse fragments over light machine gun
and mortar positions.

j This section was prepared by Walter G. Finch,
former captain in the VT detachment of the Army
Ordnance Department.

was taken from operational reports of the
Army and Navy.

962 Use by Army

The U.S. Army used these fuzes operationally
in the various theaters of operations as follows.

European Theater of Operations [ETO]

VT Fuzes, T-5. Approximately 50 T-5 fuzes
were employed by the First Tactical Air Force,
Seventh Army, during March and April 1945,
against hangars, air fields, light and heavy gun
positions. No reliable assessment data are
available because the targets were in enemy-
held territory, but informal information indi-
cated that the results obtained were good. The
general conclusion was that a larger number of
these fuzes would have been used if they had
become available about two months earlier.

VT Fuzes, T-6. None of these fuzes was used
in combat in ETO. However, rocket units of the
First Army were experimenting with the use of
these fuzes during the first months of 1945 and

OPERATIONAL USES OF BOMB AND ROCKET VT FUZES

421

had fired a total of 40 units in a demonstration
with 35 per cent random burst.

Although there was a large percentage of
random bursts, the general feeling for the fuze
was high. Following the demonstration, the
Twelfth Army Group decided that approxi-
mately one-half of all rocket projectiles used
by the ground forces should be fuzed for air
burst and they immediately placed an order for
all available T-6 fuzes. It was intended that
these fuzes be used to assist in forcing a cross-
ing of the Rhine. However, the crossing was
made ahead of schedule without much difficulty.
After this occurred, the attitude of the First
Army toward the use of rockets had cooled to
some extent, and the rocket units were dis-
banded. There were approximately 79,000 T-6
fuzes available in ETO when World War II
ended. In the type of warfare experienced in
ETO in the last phases of World War II
(a fast-moving offensive), ground-to-ground
rocket firing is not usually employed. This type
of firing is used when the front lines are stable,
and definite positions or areas are to be cap-
tured and there is a shortage of artillery weap-
ons. This was not the case in either ETO or
the Mediterranean Theater of Operations
[MTO].

VT Fuzes , T -50-El. Approximately 1,300
T-50-E1 fuzes were employed by the Ninth
Bombardment Group, Ninth Air Force, during
March and April 1945.

The initial mission was carried out on March
15, 1945 by units of the Ninth Bomber Com-
mand. Thirty-seven B-26 aircraft participated
in the attack, carrying a total of 524 260-lb
(M-81) fragmentation bombs. The target areas
were located at Pirmasens and Neunkirchen,
Germany, and consisted of important flak posi-
tions guarding avenues of approach into inner
Germany. The aircraft, flying in formations of
three and six at 15,000 ft, released the bombs
at 100-ft train spacing. Visual, verbal, and
written reports indicate that the flak was re-
duced considerably and in some cases stopped
completely.

Additional missions were against similar po-
sitions with similar results.

It was apparent that the using arms would
have used more of these fuzes in ETO if they

had been available there. Two days before the
war in ETO ended, a cable was dispatched to
the Air Ordnance Office in Washington, D. C.,
requesting immediate air shipment of 5,000
T-50-E1 fuzes to the theater.

Mediterranean Theater of
Operations [MTO]

VT Fuzes, T-5. These fuzes were not em-
ployed operationally in MTO because of Air
Force tactics. When the air forces employed
rockets for use against the enemy, they sent
their planes into combat in close proximity to
the ground. This prohibited the use of the T-5
fuzes because of range dispersion at shallow
dive angles.

VT Fuzes , T-6. The initial use of VT fuzes
(T-6) occurred during the week of March 12,
1945 when rocket units of the Fifth Army fired
the 4.5-in. rockets from ground mounts for the
first time in MTO. The target was a small ham-
let at the foot of a mountain across a deep
valley. It is estimated that 70 per cent of the
100 units fired operated normally. Rockets were
not used extensively in MTO because of the fact
that they were too erratic, and the using and
artillery arms did not have much confidence in
them because of the large range dispersion.
During March and April approximately 500
units were used in combat; at least 70 per cent
functioned satisfactorily. Deep interest was
displayed in the VT fuzes for rockets by the
using arms, and it was felt that these fuzes had
great possibilities.

VT Fuzes, T-50-E1. The Fifteenth Air Force
used approximately 1,500 T-50-E1 fuzes in
combat up to May 1, 1945. These fuzes were
employed in 260-lb fragmentation bombs
against enemy flak positions that defended an
avenue of approach into Austria and Germany.
The initial use was on April 1, 1945 when 18
aircraft of the Fifteenth Air Force dropped
213 260-lb fragmentation bombs (M-81)

against four 4-gun German flak batteries lo-
cated in six different target positions near
Grisolera, Italy.

The excellent pin-point bombing secured
many near misses on three of the four batteries
assigned. The B-24’s pilots were briefed to
attack in two waves of nine aircraft composed

422

ANALYSIS OF PERFORMANCE

of three 3-ship elements. Each element was
assigned a separate 4-gun battery. All of the
batteries attacked in the first wave ceased firing
when the bombs exploded, even though one of
the four batteries was missed by several hun-
dred yards. Fifteen minutes after the first wave
attacked, the second wave dropped their bomb
load over the same positions and reported that
all flak firing ceased as the bombs exploded.
Both waves received light, inaccurate antiair-
craft fire on their bomb runs which were made
between 24,000 and 26,000 ft. No American
planes were damaged nor were apy losses sus-
tained. Ground scores indicated that 22 soldiers
were killed, 18 were wounded, and one 20-mm
gun was destroyed.

Analysis of strike photographs taken on the
mission indicated further that (1) there were
a number of early bursts, (2) that when the
fuzes functioned properly the detonation oc-
curred approximately 17 ft off the ground, (3)
that the distribution of fragments from each
bomb over the ground was approximately
circular, and (4) that the fragments were not
uniformly dense throughout the pattern.

Later attacks under similar conditions
yielded results comparable to these of the first
mission.

VT Fuzes, T-51. The Twelfth Air Force used
approximately 100 VT fuzes, T-51 in 260-lb
fragmentation bombs, 500- and 1,000-lb GP
bombs, and the 165-gal fuel tank incendiary
bombs against enemy positions. The results of
the combat tests indicated that the fuzes could
be used to initiate 500- or 1,000-lb GP bombs
or the 260-lb fragmentation bomb could be em-
ployed successfully against personnel or equip-
ment targets that are sheltered from ground
level artillery projectiles or bomb bursts by
walls, revetments, or fox holes. The users con-
cluded that the fuzes could also be employed
effectively in carpet bombing in support of a
ground forces’ offensive.

VT Fuzes, T-51 -El. The Twelfth and Fif-
teenth Air Forces were ready to start employ-
ing the T-51-E1 fuze when it was announced
that the war had ended in MTO. There were
10,000 of these fuzes on order from the Zone of
Interior and they were scheduled for delivery
in May 1945.

Pacific Theater of Operations [POA]

VT Fuzes, T-5 and T-6. None of these fuzes
was used operationally in the Pacific War Zone.

VT Fuzes, T-50-E1 and T-50-EU . (1) The
total fuzes expended by the Army in POA until
August 1, 1945, were 1,426 T-50-E1 and 1,656
T-50-E4 fuzes.

2. A demonstration was held on January 22,
1945, at Saipan for introducing the VT fuzes
into the POA. Twelve proximity-fuzed bombs
were dropped and all the fuzes functioned prop-
erly and gave normal heights of burst.

3. The first VT fuze missions in POA were

a. February 10, 1945. Target attacked:
Air installations Iwo Jima. Ten B-24’s carried
95 500-lb GP bombs fuzed with T-50-E4 fuzes.
Results : Crews reported 65 per cent hit in the
target area. Photos showed air bursts to have
hit over a widespread area but very thinly dis-
persed except for one heavy concentration of
hits in the easternmost corner of the target
area. Several bombardiers reported that a good
percentage of the bombs exploded prematurely
(1,500 to 2,000 ft below the formation).

b. February 10, 1945. Target attacked:
AA defenses, radio and radar northeast of air
field No. 3, Iwo Jima. Results: 75 per cent of
the 50 500-lb bombs, fuzed with T-50-E4 fuzes,
hit in the target area. The fact that AA fire
ceased shortly after bombs away indicate the
possibility that this strike rendered at least
some of the AA guns inoperative.

4. Two additional missions were carried out
against Iwo Jima, several against Marcus
Island, and some at Ryukyus and Kyushu. All
attacks were either against AA installations or
airfields. Sketchy reports concerning stopping
of AA fire or early functioning of some of the
fuzes was essentially all the information re-
ceived concerning the effectiveness of the prox-
imity-fuzed bombs. Typical action photographs
are shown in Figures 32 and 33.

China, Burma, India Theaters of
Operations [CBI]

It is estimated that approximately 600 VT
fuzes were expended in CBI against flak posi-
tions and light fortifications.

VT Fuzes, T-5 0-El and T-50-EU . (1) The

OPERATIONAL USES OF BOMB AND ROCKET VT FUZES

423

Tenth Air Force expended 75 T-50-E1 fuzes
in 260-lb fragmentation bombs against AA
positions in the air preparation for landings at
Rangoon. All the fuzes operated normally.

2. The Twentieth Air Force employed 74
T-50-E1 and 349 T-50-E4 fuzes in 260-lb frag-
mentation and 500-lb GP bombs on two night
raids against flak positions and light installa-
tions. The function of the fuzes was reported
as excellent, with AA fire stopped and huge
fires started. This air force placed an order for
179,000 T-51 type fuzes.

Figure 32. Strike photograph from bomber,
illustrating fragmentation patterns obtained on
beach fortification area, Iwo Jima, February 17,
1945. Patterns are from several trains of 260-lb
fragmentation bombs, fuzed T-50-E1, released
from 5,000 ft. (Army Air Forces photograph.)

3. The Three Hundred and First Fighter
Wing employed 74 T-50-E4 fuzes on August 15,
1945 against enemy positions. Because of a
slight overcast, it was difficult to observe the
results.

VT Fuzes , T-51 -El. The Fourteenth Air
Force dropped a total of 96 T-51-E1 fuzes
against enemy AA positions, buildings of
Chinese construction, and entrenched personnel,
48 on 500-lb GP bombs, 32 on 250-lb GP bombs,
and 16 on 260-lb fragmentation bombs. In all

cases where VT-fuzed bombs were used, they
functioned properly and effectively. No mal-
functions were observed.

963 Use by Navy

The U. S. Navy employed the VT fuzes for
bombs with success against the Japanese.
These fuzes were used in missions against anti-
aircraft gun positions, light buildings, and per-
sonnel in the open. The aircraft carrier USS
Randolph, for example, employed a consider-
able number of the fuzes in the last six weeks
of World War II. From July 1 through August
15, 1945, the carrier’s aircraft dropped a total
of 2,240 bombs of all sizes over Japanese tar-
gets. Of this number, approximately 800 of the
bombs were fuzed with VT fuzes or 35 per cent
of the total number of bombs dropped during
the period were VT fuzed.

Other examples of the percentage of VT
fuzes, T-50-E1 and T-50-E4, used in combat
during the latter stages of the war with the
Japanese are listed below.

From July 10 to August 15, 1945

T-50

Conven-

fuze,

type

tional fuze,

260-lb

500-lb

all types

Aircraft carrier

frag.

of bombs

% VT

USS Bennington

348

170

1,204

USS Independence

242

431

USS San Jacinto

212

712

USS Shangri-La

584

185

1,300

Reports on Effectiveness. Following are ex-
tracts from various Navy reports on the use of
VT bomb fuzes.

1. Excerpt from Report USS Yorktown for
the period from May 24 to June 13, 1945, sup-
port of Okinawa operations:

VT fuzes were used with both 260-lb fragmentation
and 500-lb GP bombs, this ship’s first experience with
these fuzes. Pilot observations as to fuze functioning
and bomb effectiveness were necessarily limited because
of the high release altitudes required with these fuzes
and the type damage done by fragmentation bombs, but
the pilots were generally enthusiastic about the possi-
bilities of this type of attack. The latest VT fuzes,
which have reasonably low minimum release altitudes,
in fragmentation bombs promise to be excellent weapons
for use against revetted aircraft and personnel targets.

424

ANALYSIS OF PERFORMANCE

2. Excerpt from brief of Commander Task
Group 38.4, dated May 24 to June 13, 1945:

VT fuzes were employed for the first time during this
operation. Functioning of the fuzes appeared satis-
factory, but an accurate count could not be obtained.
The best available information indicates about ten per
cent were duds, exploding on impact and another ten
per cent exploded prematurely. Some of the prematures
were possibly caused by close proximity to other bombs.
The high release altitude required to arm these fuzes is
a distinct disadvantage. Fuzes requiring shorter travel

saturation of defenses were achieved by having all
available VF and VBF strike a single airfield system
in a coordinated plan over the shortest possible time
interval. Ample time was allowed for careful target
assignment and briefing. An approach track which
allowed the enemy minimum warning was selected.
Finally, a weapon was selected (260-lb fragmentation
bombs, VT-fuzed) which apparently effectively attacked
revetted aircraft and anti-aircraft positions. This opera-
tion was entirely successful; considerable damage is
estimated to have been done the enemy with the loss to
ourselves of no pilots and only four airplanes.

Figure 33. Strike photograph from bomber, illustrating fragmentation patterns obtained on air field
at Tsuiki, northern Kyushu, August 8, 1945. Bombs were 260-lb fragmentation, fuzed T-50-E1, released
from 10,000 ft. Some bombs burst over water, giving sharply defined fragmentation patterns (Army Air
Forces photograph) .

to arm should be made available as soon as possible.
VT fuzes are a valuable addition to our offensive arma-
ment, but it is felt that strafing is still the primary
means of destroying revetted aircraft.

3. Extract from a Task Force 38 report,
dated June 8, 1945 :

This operation is given separate treatment because it
was specifically planned to avoid the difficulties of the
previous Kyushu sweeps. Tactical concentration and

4. Aircraft launched from the USS Ticon-
deroga on June 9 and 10, 1945, were used to
drop 260-lb fragmentation bombs fuzed with
T-50-E1 fuzes and 500-lb GP bombs fuzed with
T-50-E4 fuzes on antiaircraft positions on
Minami Shima and Kita Shima. The pilots of
the aircraft estimated that 90 per cent of the
106 VT-fuzed bombs dropped functioned nor-
mally and that the antiaircraft fire from the

OPERATIONAL USES OF BOMB AND ROCKET VT FUZES

425

islands, in general, ceased after the attack.

5. Excerpt from brief of action reports and
analysis of strike on Wake Island, June 20,
1945, ComCarDiv 11, USS Hancock , USS Lex-
ington, and USS Cowmens:

Fighters were used exclusively on anti-aircraft and
several installations appear to have been knocked out.
Favorable reports were made on the effectiveness of the
air burst (VT) fuze by Air Group SIX. This group
employed their VT bombs in what appears to be a most
effective manner. The first dive was made for the sole
purpose of releasing the air burst fuzed bombs, the
second pass, using their rockets and machine guns, led
the bombers in to the target. It is interesting to note
that none of the Hancock bombers were hit. While the
white phosphorus bombs seem to have functioned
normally, it is believed that a certain amount of train-
ing is required by the pilots carrying this bomb to
provide practice in placing the bomb properly with
relation to the target, and by bombers who must learn
how to wait until the smoke cloud has had time to
develop fully before coming within AA range.

The attacks at Wake were characterized by more
extensive anti-flak measures than naval A/C have per-
haps ever used from the point of view of ordnance and
tactics. When the strike group sighted the island at
about 30 miles, anti-flak VF broke away, flew in ahead
and attacked threatening AA positions with VT fuzed
bombs (usually 260-lb frags) and White Phosphorus
Bombs. VF then rejoined the group orbiting 10 to 15
miles away and a coordinated attack followed with VF
rocketing and strafing AA positions a few seconds
ahead of VB, VT, and VBF.

260-lb Frag M-81 with VT Fuzing: Reports of
observers indicate that this bomb and fuzing may be
very effective. Several AA installations, medium and
heavy, were definitely silenced, but whether this can be
attributed to personnel or material casualties cannot
be determined at this time. Photographs do not reveal
definite material damage, although it may be exten-
sive.

1,000-lb GP Bombs with VT Fuzes: An experienced
ACI officer observer believes that use of this bomb must
have caused extensive damage, although again it is not
revealed by photographs. Bursts were just above ground
and high enough to clear revetments.

VT Fuzes T -5 0-El and T-50-EU: This fuze is very
effective and must be carefully considered in planning
bomb loading. One dud was reported although all VT-
fuzed bombs also had tail fuzes. No prematures were
reported. The relatively high point of release for arming
is a disadvantage in pinpoint bombing, but the T-90
series VT fuzes should tend to overcome this defect. The
results of this operation indicate that VT-fuzed bombs
should be highly effective against heavily revetted posi-
tions, anti-aircraft positions, personnel, parked aircraft,
and vehicles. If pilots were experienced in train release
of VT-fuzed bombs, more practical loadings could be

made. In this operation only one VT-fuzed bomb per
plane was used.

It is believed that the use of the VT-fuzed bombs by
the anti-flak fighter planes of the Air Group was highly
successful against anti-aircraft positions attacked on
Wilkes Island and Wake Island. Two medium A A near
the Marine Camp of Southwest Wake Island were
permanently silenced after a VT bomb attack and some
of the many guns at and near Peacock Point may have
been knocked out. VB and VTB encountered almost no
AA fire on their attacks although attacking VF were
subject to fire from heavy, medium, and light AA. It is
reasonable to believe that the VB, VRB, immunity was
due to the anti-flak attacks preceding the bombing runs.
Commander of VBF 6 reported that the VT-fuzed bomb
burst left wide circular residual smoke on the ground
estimated at least 300 ft in radius. Plainly the aerial
bursts with their wide-spread fragmentation and blast
damage may have inflicted substantial casualties to
personnel, aside from their psychological and morale
effects.

Some uneasiness was experienced by pilots in using
the VT-fuzed bombs at the possibility of the arming
wire slipping out and the bomb being armed by its air
travel while still hung on the wing rack.

The T-50-E1 and T-50-E4 fuzes require too high
minimum-release altitudes for accurate bombing. Issue
of the newer T-91 and T-92 fuzes, when available, will
materially increase the effectiveness of the VT bombs.

This Air Group dropped a total of 42 bombs with
VT aerial burst bomb fuzes. Very little tangible results
could be observed from the use of these bombs. In some
instances pilots observed slight aerial disturbances
over targets where bombs were dropped with slight
dust clouds and other debris. Since no personnel were
observed, the effect of these bomb bursts against per-
sonnel could not be ascertained.

As target coordinator on certain strikes, I observed
no tangible evidence of the effect of the bombs bursting
in the air except the slight dust disturbances. No
diminishing of AA fire can be definitely attributed to
these bombs. Some of these VT-fuzed bombs were seen
to explode by contact. However, due to pilots inability
to observe bomb explosions during dives, there may
have been many more bombs exploded by contact.

AirPacComment : The “slight dust disturbances”
mentioned arise from impact of fragments on the
ground about the burst. The presence of such a dust
pattern, and of an orange explosion and black smoke
billowing in all directions, is evidence of an aerial burst.
Damage visible from the air will seldom be inflicted by
VT-fuzed bombs, but the extent of the dust pattern will
show the extent of the pattern of lethal and damaging
fragments. The concussion of a nearby aerial burst, par-
ticularly of a large GP, is also likely to be somewhat
unsettling to AA gunners.

6. Excerpt from CO USS Coivpen’s report,
dated June 23, 1945 :

426

ANALYSIS OF PERFORMANCE

The effectiveness of subject bombs (VT fuzed) was
difficult to observe from the air. Pilots were at times
unable to judge whether the bombs burst on or above
the ground, but the consensus of opinion is that the
majority of bursts were above ground level.

Two observed cases of reduction of AA fire after
attack with subject bombs were noted as follows:

(a) Medium AA fire from the vicinity of . . . ap-
parently ceased after strikes one and two.

(b) Heavy A A positions at . . . were attacked, and
at least two very close air bursts were obtained. Heavy
AA guns were observed firing from this position imme-
diately prior to attack. Five VF aircraft of VF-50
made a second dive on this position five minutes later,
and all pilots stated definitely that the guns were not
firing on the second attack.

In the opinion of the Commander, . . . , the 260-lb
fragmentation bomb with VT fuze is an excellent
weapon for attacks on revetted positions and is far
superior to WP bombs and to rockets. It is recommended
that two such bombs per VF aircraft would make an
excellent load for all attacks on AA positions, personnel,
grounded aircraft or vehicles. When the 260-lb fragmen-
tation bomb becomes available, it should be even better
for this purpose.

Due to the fact that salvo drops of VT-fuzed bombs
are inadvisable, it is further recommended that, if
practicable, VF bomb releases be rewired through the
rocket selector box so that drops in train may be made
more easily.

7. Excerpt from action report USS Essex ,
July 2 to August 15, 1945:

VT Prematures: No accurate statement can be made
of the number of VT prematures dropped by bombers,
but it is estimated to be below 10 per cent. The majority
of prematures appear to have been dropped by fighters
who released higher than bombers. It is estimated that
the number of prematures from fighters was sometimes
over 50 per cent. The only prominent variables involved
were that fighters carry these bombs externally and
that they be released at speeds 70 to 100 knots higher
than the bombers. It is suggested that experiments be
coffaucted to determine whether speed at release has
any effect on premature bursts. It was the opinion of
most of the pilots that the premature bursts were 500-lb
GP bombs rather than 260-lb frags. There can be no
certainty about this observation, since judgment could
be made only from the appearance of the burst, but it is
an indication that the fuze used with the 500-lb GP is
more susceptible to premature functioning than the one
used with the 260-lb frag.

VT-Fuzed Bombs: Greatly increased damage per ton
of bombs dropped on revetted and parked airplanes is
believed to result from the use of T-50-E1 VT-fuzed
frag bombs. However, somewhat less enthusiasm is felt
for the T-50-E4 VT-fuzed 500-lb GP bombs. In the
case of the latter, a considerably higher percentage of
“prematures” is indicated from the overall evidence

that is available. Further, the actual total damage per
load of bombs is believed to be greater (assuming 100
per cent correct fuze performance) in the case of the
260-lb frag bombs. Also, the rather strenuous ordeal of
the TBM to attain a climbing or cruising speed high
enough to satisfy the SB2C’s, F6F’s, and F4U’s at
16,000 to 20,000 ft altitude when loaded with 4 x 500-lb
bombs, brings to issue a point in favor of the lighter
load of 6 x 260-lb bombs.

8. Excerpts from CNO, memo, dated July
14, 1945:

VT-Fuzed Bombs: On the Kanoya strike 8 June, VT
fuzes were used for the first time. These were attached
to all bombs released over the target area (52 260-lb
Frag and 11 500-lb GP). Pilots observed a few pre-
mature bursts, but the general opinion was that func-
tioning of these fuzes was satisfactory and that the
area was well covered with bursts exploding close to
and above ground.

9. Excerpt from Commander Air Force,
Pacific Fleet on Japan Operations July 10 to 18,
1945 :

VT-fuzed bombs used extensively against parked A/C
in the Tokyo district are believed to be an ideal loading
for this type of target. Some high bursts were observed,
but the number of these was less than the anticipated
10 per cent. The required high release altitudes in
reducing bombing accuracy emphasized the importance
of issuance of the new T-91 and T-92 fuzes.

The VT fuzes referred to were of the T-50 type.

10. Excerpt from USS Lexington action re-
port, dated August 4, 1945 :

This Air Group dropped a total of forty-two (42)
bombs with VT aerial burst bomb fuzes. Very little
tangible results could be observed from the use of these
bombs. In some instances pilots observed slight aerial
disturbances over targets where these bombs were
dropped with slight dust clouds and other debris. Since
no personnel were observed, the effect of these bomb
bursts against personnel could not be ascertained.

The target coordination observed no tangible evidence
of the effect of the bombs bursting in the air except the
slight dust disturbances. No diminishing of AA fire can
be definitely attributed to these bombs. A total of three
VT-fuzed bombs were seen to explode by contact. How-
ever, due to pilots inability to observe bomb explosions
during dives, there may have been more bombs exploded
by contact.

11. Excerpt from action report, USS Ben-
nington, dated August 31, 1945; operations
against the Japanese homeland from Western
Honshu to Eastern Hokkaido:

VT Fuzes. Although positive damage assessment is
extremely difficult, it is believed that VT fuzes have

OPERATIONAL USES OF BOMB AND ROCKET VT FUZES

427

performed extremely well and that they have solved the
long-present air burst fuze problem. If all safety pre-
cautions are strictly followed, they are as safe as the
conventional fuzes and no trouble whatsoever will be
encountered.

9'6'4 Summary of Conclusions Made by
Using Arms

An analysis of operational reports by the
services yielded the following general conclu-
sions :

1. The general attitude of the using arms to
the bomb and rocket VT fuzes at the end of
World War II was most favorable. Originally
there was much doubt as to their possible value
as a lethal weapon. The general attitude was
that the fuzes had very limited use, that they
were unsafe, and that a high percentage of
them malfunctioned. Combat experience in the
various theaters changed this view, and with
the close of World War II the using arms were
very enthusiastic over the fuzes.

2. The operational use of VT bomb and
rocket fuzes, particularly in ETO and MTO,
was retarded by transmission to the theaters
(by the AAF) of unfavorable data taken at
Eglin Field on preproduction fuzes. This infor-
mation caused a feeling that the fuzes were
not ready for operational use at that time and
necessitated a great deal of experimenting in
the combat area to see if they performed satis-
factorily and were safe. Exhaustive preopera-
tional tests were conducted in both ETO and
MTO before employing the fuzes operation-
ally.

3. These fuzes could be used effectively to
explode 260-lb fragmentation bombs or 500- or
1,000-lb GP bombs on personnel or equipment
targets that were sheltered from ground level
artillery projectiles or bomb bursts by walls,
revetments, or fox holes.

4. These fuzes could be used effectively in
carpet bombing in close support of a ground
force offensive. It was felt that an area could
be saturated with bomb fragments at an angle
that is most effective against personnel occupy-
ing defensive positions.

5. These fuzes could be used effectively for
neutralizing concentrated flak positions, such

as were found on many of the Pacific islands.
Bombers at high altitudes can identify flak
positions and drop 260-lb fragmentation bombs
with VT fuzes accurately enough to cause a
diminution of accuracy and intensity of AA op-
position.

6. These fuzes could be used effectively with
500- and 1,000-lb GP bombs mounted under the
wings of fighter aircraft.

7. The T-5 and T-6 type fuzes had very lim-
ited use because of our superiority in airpower
and the high dispersion of the M-8 rocket.

8. The T-50-E1 and T-50-E4 type fuzes gave
an operational performance of 70 to 75 per cent
in actual combat, based on reports submitted
from the various theaters. These figures are
lower than test results of 80 to 85 per cent
obtained in the United States, and this is prob-
ably due either to incomplete counts or to poor
installation of the fuzes in the combat areas.

9. Ground reports from special agents indi-
cate that fragmentation bombs with VT fuzes
reduce morale and accuracy of flak personnel,
kill and injure flak personnel, and cause dam-
age to AA equipment, such as cables, directors,
and radar units.

10. In attacking flak positions, it was found
by the Ninth and Fifteenth Air Forces that the
best tactics were to attack each position in ele-
ments of threes instead of a large number of
planes in close formation.

11. In addition a number of suggestions
were made for improving or modifying the
fuzes. These included

a. That the percentage of random bursts,
especially the early bursts, be reduced both for
psychological and economic reasons.

b. That arming delays be supplied with
all bomb fuzes being shipped in order to im-
prove fuze performance and give an added
margin of safety whenever possible.

c. That provision be made for shorter
arming times, particularly for dive bombing.

d. That extra lock washers be supplied
with shipments of the fuzes, since they are usu-
ally flattened during installation or removal of
the fuzes.

e. That a streamlined windshield be de-
veloped to cover the arming vane for those
fuzes that are employed in bombs that are

428

ANALYSIS OF PERFORMANCE

mounted under the wing racks in fighter type
aircraft.

f. That corrective measures be insti-
tuted to eliminate breaking of cotter pins on
arming vanes and thus causing duds.

g. That a suitable wrench be shipped
with the fuzes for tightening the fin-locking
nuts on the various bombs.

h. That improvements be made in the
arming wire method of preventing vane rota-
tion, with particular attention to bombs car-
ried under the wings of fighter planes. Al-
though the arrangement appeared satisfactory
when properly installed, an error in installation
would allow the arming wire to be pulled out by
air drag while still mounted on the wing.

9 7 SUMMARY AND CONCLUSIONS

Pertinent summary data of this volume on
radio proximity fuzes were presented in the
introduction, particularly in Sections 1.4 and
1.5. The reader may therefore refer to Chapter
1 for summary information.

With proximity fuzes established as impor-
tant and practicable ordnance items, it is desir-
able that development work continue. In the
various preceding chapters, as well as in this
chapter, the limitations and deficiencies of the
fuzes developed during World War II have
been discussed. Future work will naturally at-
tempt to eliminate these deficiencies. It has
been pointed out in several places in the vol-
ume that numerous compromises in design
were necessary for reasons of expediency. In
an orderly long-term peacetime development,
such compromises should be less difficult to re-
solve.

A detailed discussion of the limitations of the
fuzes previously described and methods for
improvement should not, however, be made
here, for two important reasons.

1. Advanced thinking and more sophisti-
cated development on radio proximity fuzes
will be classified “secret” much longer, accord-
ing to present classification policy, than will the
material presented in this volume. Accordingly,
the possible circulation of this volume would be
appreciably curtailed by including, just in sug-

gestive form, some of the most promising ideas
for fuze improvements. The material in this
volume has been presented, as far as possible,
in a form to provide a basis for further devel-
opment. The subject matter of Chapter 2, in
particular, is fundamental to any fuze design
which involves the interaction of radio waves
with a target. Thus, it does not seem desirable
to impair the possible usefulness of this volume
by including a little new and much more re-
stricted material which is of unproven merit.

2. One of the main reasons for outlining sug-
gestions for possible future work is to urge that
such work be undertaken. The Army Ordnance
Department, one of the agencies to whom this
report is made, has already formulated a vig-
orous fundamental development program on
proximity fuzes. Assumption of responsibility
for further development was started by the
Army prior to the end of World War II and is
continuing. Division 4’s central laboratories,
the Ordnance Development Division at the Na-
tional Bureau of Standards, are now working
for the Army Ordnance Department on new
fuze problems. Thus, with an active far-reach-
ing program on proximity fuzes already under
way, it becomes superfluous to suggest here
what form that program might take.

The important thing is to insure that the ac-
cumulated technical information and experience
of World War II period is available in an orderly
form for those who will continue the work. It
is hoped that the preceding pages of this vol-
ume have fulfilled that objective.

98 APPENDIX TO CHAPTER 9

ACCEPTANCE TEST CONDITIONS

981 Acceptance Testing of Bomb Fuzes

A program for acceptance testing of VT
bomb fuzes was established early in 1944. Since
then minor variations in acceptance require-
ments have been made. Basically the testing
procedure has remained unchanged. Every
manufacturer’s lot, considered for acceptance,
was subjected to two types of field tests : a metal
parts assembly test, followed after acceptance
by a loading test of ammunition lots, which

APPENDIX TO CHAPTER 9

429

usually involved several metal parts lots. An
outline of Army Ordnance specifications for
these tests follows :51

A. Metal Parts Test

1. A ballistic sample of 18 metal parts assemblies
(fuzes) prepared for testing will be shipped to the
Proving Ground from a loading plant.

2. Metal Parts to be tested will be assembled to
bombs, as follows :

a. T-50-E1, T-89— Bomb, Frag, 260 lb,
AN-M81.

b. T-50-E4, T-90— Bomb, GP, 500 lb, AN-M64.

c. T-51, T-51-E1— Bomb, Frag, 260 lb,
AN-M81 or Bomb, GP, 250 lb, AN-M57.

d. T-91— Bomb, Frag, 260 lb, AN-M81.

e. T-92— Bomb, GP, 500 lb, AN-M64.

The AN-M64 and AN-M57 will be sand loaded. The
AN-M81 may be used empty (empty-weight 220 lb).
The bombs should be equipped with a suitable spotting
charge. Every precaution will be taken that both the
fuze and tail fin assembly are tightly screwed to the
bomb. (Not possible to unscrew by hand. A wrench is
provided to tighten the fuze.)

3. All bombs will be dropped singly in the normal
manner from an aircraft flying at a true air speed of
200 ± 5 mph from a true altitude of 10,000 — 1,000 ft
above the target.

4. Normal Test Plan.

a. Seventeen metal parts assemblies will be
tested for a first sample.

b. Requirement for Acceptance — 12 or more
assemblies shall cause proper functioning.

c. Retest sample will contain 23 metal parts
assemblies.

d. Requirement for Acceptance on Retest — 26
or more assemblies.

5. Reduced Test Plan. (This plan was discon-
tinued April 3, 1945 by order of Army Ordnance.)

a. Six metal parts assemblies will be tested
for the first sample.

b. Requirement for Acceptance — 5 or more
assemblies must cause proper functioning.

c. Twelve metal parts assemblies will con-
stitute a retest sample.

d. Requirement for Acceptance on Retest — a
total of 10 or more assemblies out of the entire 18
tested shall cause proper functioning.

6. The following procedure applies to the Reduced
Testing Plan of Paragraph 5 above:

a. If 10 successively produced lots offered for
acceptance by a manufacturer are accepted under the
Normal Test Plan the producer is placed on a preferred
list which entitles him to have his product tested on the
basis of the reduced testing criterion.

b. Such a manufacturer will remain on this
basis until a lot is rejected on the basis of the Reduced
Test Plan. The producer will then return to the Normal
Testing Plan basis.

c. Requalifi^ation as explained in (a) above

will be necessary in order for the manufacturers
product to be again tested on the Reduced Testing Plan.

7. Height of burst for proper function will de-
pend on the nature of the target area. If the metal parts
assemblies are tested over water the height of burst for
proper functions are fixed and are listed below. If the as-
semblies are tested over land the required heights for
proper function shall be determined by multiplying the
required height range for testing over water by a target
factor which is to be determined at least twice during
the testing period of a day. This factor depends upon
variables of the target area. Necessary equipment to
obtain this information and personnel to install and
instruct in its use will be arranged through the Office,
Chief of Ordnance. The following figures constitute the
range of proper functioning when the fuzes are tested
over water:

T-50-E4, T-50-E1, T-89, T-90, T-91 and T-92-
between 10 and 160 ft.

T-51, T-51-E1 — between 60 and 240 ft.

B. Loading Tests

1. Accepted metal parts lots will be received at
the loading plant and be assembled into a grand lot for
loading. Lot for loading will generally consist of more
than one metal parts lot of one manufacturer. From
each loaded lot a ballistic sample of 20 fuzes will be
shipped to the Proving Ground for test.

2. The sample sent to the Proving Ground will be
tested for the following qualities :

a. Minimum Safe Air Travel (abbreviated
MinSAT).

b. Functioning quality of the loading com-
ponents.

3. The Loading tests will consist of 2 phases.
Phase 1 is a test of the MinSAT of the lot and con-
ducted with inert bombs ; phase 2 is a test of functioning
quality of the lot and is conducted with HE bombs.
Phase 1 is based upon an altitude of release from which
no arming should occur before impact. Phase 2 is based
upon an altitude of release from which the majority of
the fuzes should arm before impact.

4. Arming of VT bomb fuzes is dependent upon
the size and shape of the bomb as well as its ballistic
character of flight; therefore, it is desirable to conduct
phase 1 of the test on the bomb which will give the fuze
the least MinSAT with which that particular model of
fuze may be used. At present this bomb is the AN-M30
or the AN-M81 for all models.

5. Below are tabulated the requirements for con-
ducting Loaded Acceptance Tests of VT bomb fuzes.

6. Phase 1 may be released in train at any desired
interval and should impact on normal soil. Phase 2, if
desired, may be released in train providing AN-M30 and
AN-M81 bombs are spaced with at least 50 ft. interval
and AN-M57 and AN-M64 are spaced with at least 100
ft. interval. Phase 2 may be tested over water or land.
Phase 2 should not be conducted if a lot fails on phase 1.

7. A lot of fuzes should be rejected if it fails to
meet the requirements of either phase 1 or 2 of this test.

430

ANALYSIS OF PERFORMANCE

Upon rejection, a retest will be authorized only by the
Office, Chief of Ordnance.

98 2 Acceptance Tests of Navy Rocket
Fuzes

A procedure similar to that used for bomb
fuzes was followed in the acceptance testing of

the T-2004 rocket fuze. The major part of the
rocket testing program was done in accordance
with Army Ordnance specifications of May
1945. 52* 53 A summary of these specifications is
given below.

A lot was tested in two phases : first a metal
parts test, and then, provided the first had been
passed, a loading acceptance test.

Fuze, Bomb,

Nose, VT, T-50-E1 & T-89,

3,600 ft MinSAT (Sample Containing 20)

True air-

True alti-

Requirements

Phase

Quantity

Bomb

speed (mph)*

tude (ft)*

for acceptance

fAN-M81 inert (empty)

200

1,750 — 200

10 duds

with spotting charge

, ,

AN-M30 HE

200

3,200 + 200 (

8 or more high-

AN-M81 HE

200

3,200 + 200 \

order functions

Fuze, Bomb,

Nose, VT, T-50-E4 & T-90,

3,600 ft MinSAT (Sample Containing 20)

True air-

True alti-

Requirements

Phase

Quantity

Bomb

speed (mph)*

tude (ft)*

for acceptance

fAN-M81 inert (empty)

200

1,750 — 200

10 duds

with spotting charge

AN-M30 HE

200

3,200 + 200 )

8 or more high-

AN-M64 HE

200

4,100 + 200 S

order functions

Fuze, Bomb,

Nose, VT, T-51 or T-51-E1,

3,600 ft MinSAT (Sample Containing 20)

True air-

True alti-

Requirements

Phase

Quantity

Bomb

speed (mph)*

tude (ft)*

for acceptance

AN-M81 (empty) with
spotting charge

200

1,700 — 200

10 duds

8 or more high-

AN-M57 HE

200

3,400 + 200

order functions

Fuze, Bomb,

Nose, VT, T-51 or T-51-E1,

4,500 ft MinSAT (Sample Containing 20)

True air-

True alti-

Requirements

Phase

Quantity

Bomb

speed (mph)*

tude (ft)*

for acceptance

AN-M81 (empty) with
spotting charge

200

2,400 — 200

10 duds

8 or more high-

AN-M57 HE

200

4,500 + 200

order functions

Fuze, Bomb, Nose, VT, T-91, 2,000 ft MinSAT (Sample Containing 20)

True air-

True alti-

Requirements

Phase

Quantity

Bomb

speed (mph) *

tude (ft)*

for acceptance

fAN-M81 inert (empty)
with spotting charge

200

600 + 100

10 duds

AN-M30 HE loaded

200

3,900 + 200

4 or more high-

with Arming Delay, Air
Travel T2E1 set at 3
Divisions

order functions

AN-M81 (empty) with

200

1,350 + 200

4 or more func-

spotting charge

tions

Fuze, Bomb, Nose, VT, T-92, 2,600

ft MinSAT (Sample Containing 20)

True air-

True alti-

Requirements

Phase

Quantity

Bomb

speed (mph) *

tude (ft) *

for acceptance

•|AN-M81 (empty) with
spotting charge

200

900 — 200

10 duds

AN-M30 HE loaded

200

3,900 + 200

4 or more high-

with Arming Delay, Air
Travel T2E1, Dial Set-
ting at 3 Divisions

order functions

AN-M64 empty (sand-

200

2,400 + 200

4 or more func-

loaded to weight) with
spotting charge

tions

* The true altitude and true airspeed at release should be carefully adjusted as the results of this test depend nearly entirely upon them.

Only level flight, true altitude and true airspeeds as listed will give the

correct results.

t If AN-M81 empty bombs are

not available, AN-M30 sand-loaded

bombs may be used.

SECRET*.

APPENDIX TO CHAPTER 9

431

Metal Parts Test. In general, 17 fuzes from
each manufacturer’s lot tested were fired over
water on 3.25-in. Mk-7 motors with empty
3.5-in. Mk-5 heads. The rockets were launched
singly from a ground rail installation with a
firing elevation of approximately 30 degrees.

Fuzes from every tenth lot, however, were
fired from a plane over a water target at any
convenient dive angle not less than 20 degrees.
Plane speeds were approximately 250 mph.

Requirements for acceptance: Twelve or
more units were required to function properly
on approach to water. (Mid-flight functions
after 6 sec of flight time were considered proper
functions.) Proper function height limits were
10 to 100 ft. No function was to occur before
450 ft of air travel.

In case of failure a retest of 23 additional
units was made. Twenty-six of the total of 40
units tested were required to function as indi-
cated above.

Loading Test. Lots which had passed the
metal parts test were loaded at Picatinny Ar-
senal and combined into larger lots. These were
subjected to a mechanical arming test and to a
functioning test.

Mechanical arming test: Ten fuzes, wired to
function on mechanical arming, were fired on
3.25-in. Mk-7 motors with 5.0-in. Mk-1 HE
heads. The rockets were launched singly from a
ground rail installation at any convenient eleva-
tion.

All fuzes were required to function between
290 and 650 ft of air travel. The failure of more
than one fuze to cause the rocket to burst with
a high-order detonation caused rejection of a
lot.

If a lot failed, a retest of 20 additional fuzes
could be made at the request of the contractor,
provided (1) no burst had occurred before 290
ft, (2) not more than one burst had occurred
after 650 ft, and (3) not more than two fuzes
had failed to cause the rocket to burst with a
high-order detonation.

For acceptance of a retested lot, it was re-
quired that on the basis of the total of 30 fuzes
tested (1) no burst occur before 290 ft, (2) not
more than two bursts occur after 650 ft, and
(3) not more than three fuzes fail to cause the
rocket to burst with high-order detonation.

Functioning test: Ten fuzes, set for normal
functioning, were fired over water on 3.25-in.
Mk-7 motors with 3.5-in. empty Mk-5 heads.
Rockets were launched singly from a ground

rail installation at an elevation of approxi-
mately 30 degrees.

A lot was accepted if no fuze functioned be-
fore 450 ft of air travel and not more than two
fuzes failed to function.

If a lot failed, a retest of 20 additional fuzes
could be made at the request of the contractor.
The retested lot was accepted provided on the
basis of 30 fuzes tested (1) no burst occurred
before 450 ft, and (2) not more than four fuzes
failed to function.

A number of lots tested under this program
were accepted despite the fact that they failed
to meet the requirements of the mechanical
arming test. It was later found that, with high
ambient temperatures, the air travel was in-
sufficient for the 100 propeller turns under ac-
celeration necessary for the first stage of the
arming process. In August 1945, revisions were
made in the original specifications.54* 55 All lots
which were retested under the new specifica-
tions passed. The highlights of the changes
made follow.

Metal parts test

1. Testing from aircraft was eliminated.

2. Proper function limits required for accept-
ance were made 10 to 70 ft, with an average
height of about 35 ft.

Mechanical arming test

1. Projectile: 3.25-in. Mk-7 motor with 3.5-
in. Mk-5 head.

2. All functions were required to occur be-
tween 300 and 850 ft of air travel.

3. Retest requirements were changed in ac-
cordance with these new limits.

Loading -functioning test

1. Projectile: 3.25-in. Mk-7 motor with 5.0-
in. Mk-1 head (inert loaded to 48 lb).

2. Rockets were to be fired over ground from
a plane at any convenient dive angle less than
20 degrees, and at an altitude to give a mini-
mum flight time of 4 sec. Plane speed: 275 mph
approximately.

3. Requirements for acceptance: No fuze
should function before 1.9 sec flight time; not
more than two fuzes should fail to function.

4. Corresponding changes were made in re-
test conditions.

9 8 3 Acceptance Testing of T-5 Fuzes

From February 1943 to May 1944, acceptance
tests were made on 365 lots of T-5 fuzes. For
the first six months, the tests were conducted

432

ANALYSIS OF PERFORMANCE

by NBS at Fort Fisher and Blossom Point. The
mock-plane targets and firing towers used at
these proving grounds are described in Chap-
ter 8. Later testing was conducted by Army
Ordnance at Aberdeen Proving Ground. A rec-
tangular wire mesh screen, stretched between
four poles, was used as the target there.

The acceptance requirements at all three
proving grounds were essentially the same.
Salient features are given below.

Test Procedure. Twenty units, from each lot
of 1,000 to be tested, were mounted on Revere
or Budd 4V2-in. rockets, and fired horizontally
from a tower for function on approach to a tar-
get approximately 70 ft above ground.

Requirements for Acceptance. At least ten
units were required to function properly. In a
considerable part of the testing, firing was
stopped as soon as ten proper functions were
obtained.

Method of Scoring. In order that a unit be
counted in scoring, it had to pass within the
radius of action of the target. For the mock-
plane targets used at Fort Fisher and Blossom
Point, the scoring region was defined by a circle
of 60-ft radius, cut off by a plane 40 ft above
the ground (because of the possibility of ground
firing) .

Functions were classified as proper, early,
late, or dud. A proper function was one occur-
ring not more than 60 ft before the center of
the wing of the mock-plane target, and not later
than 35 ft after. Functions occurring before
and after the proper-function limits were clas-
sified as earlies and lates respectively. A dud
was a unit which failed to function.

Retest conditions: If a lot failed the normal
test, 60 additional units were fired. At least 44
proper functions, based on the total of 80 units
tested, were necessary for acceptance.

GLOSSARY"

A. Approach. Representing fuze function on approach
to ground target.

A Voltage. The voltage applied to the filaments of
vacuum tubes.

A Winding. The winding or coil on the generator
power supply which furnishes A voltage.

Afterburning. Afterburning is burning in the rocket
motor occurring after the main burning or accelera-
tion period. See Sections 9.2 and 9.3.

Amplifier. That part of a radio proximity fuze which
amplifies the doppler signal to a magnitude sufficient
to fire a thyratron.

Amplifier Gain. Ratio of amplifier output voltage to
amplifier input voltage.

Antenna. The radiator or exciting portion of a radio
proximity fuze; the bars in a transversely excited
fuze, the ring in a longitudinally excited fuze, the
conical cap in the T-5 fuze, or the loop in the T-172.

Antenna Gain. Ratio of the power transmitted per
unit area in a given direction relative to that from
an isotropic antenna having the same total radiated
power. See Section 2.8.

Antenna Reactance. The reactance occurring in the
parallel resistance-reactance combination which is
equivalent to the antenna. See Section 2.7.

Antenna Resistance. The resistance occurring in the
parallel resistance-reactance combination which is
equivalent to the antenna. See Section 2.7.

Approach Angle. The angle between the trajectory
of a missile and the vertical on approach to the
ground.

Arming. Removal of the mechanical and electrical
barriers to the operation of the explosive train in a
fuze prior to which an activating signal in the fuze
cannot cause detonation.

Arming Angle. The angle through which the de-
tonator rotor turns to complete the arming cycle.

Arming Delay. The time delay between launching of
the missile to completion of arming. The term some-
times applied to the T-2 delayed arming device. See
Section 4.2.

Arming Pulse. An electrical disturbance sometimes
arising when the detonator circuit is closed.

Arming Wire. A wire attached to an aircraft which
prevents initiation of the arming cycle until the bomb
or rocket has left the aircraft.

Audio. An adjective applied to electrical circuits or
appropriate signals having frequencies in the audible
range. It usually refers to the detected doppler signal.

B Voltage. The supply voltage for the anodes of the
electronic tubes.

B Winding. The high voltage winding of the gen-
erator.

Bar Type. A transversely excited radio proximity
fuze using a center-fed transverse bar as an antenna.

a Many of the terms included in this glossary may not occur in

the text but occur frequently in the references given in the

Bibliography.

BRLG. (Bomb, Radio, Longitudinal, Generator.) An
early designation for a generator-powered ring-type
bomb fuze.

Brown. A code term for 75 megacycles per second.

BRTB. (Bomb, Radio, Transverse, Battery.) An early
designation for a battery-powered bomb fuze with
transverse antenna.

BRTG. (Bomb, Radio, Transverse, Generatpr.) An
early designation for a generator-powered radio
proximity fuze with transverse antenna.

BTL. Bell Telephone Laboratories, Inc.

Burst Surface. A hypothetical surface surrounding
the target, showing the locus of fuze bursts upon
approach to the target.

C. Capacitance.

C Bias or Voltage. The supply voltage for biasing
the pentode and thyratron.

C or Cv. An abbreviation for C voltage.

Camera Obscura. The usual camera obscura tech-
nique applied to observation of fuze function.

Cap. The ring or conical cap used as an exciting
antenna.

Carrier. The radio frequency signal generated by the
fuze oscillator.

CCM. Counter-countermeasures.

CF. Carrier frequency.

CIMA. Carrier indication of mechanical arming. See
Section 8.3.

CM. Countermeasures.

Compensated Load or Resistor. A dummy antenna
resistor with inductance to simulate the actual
antenna load.

Corncake. The proving ground at Fort Fisher.

Critical Grid Voltage. The maximum bias (negative)
at which the thyratron will fire.

D. Dud.

Delayed Arming Device. An auxiliary wind-driven
delay mechanism which locks the windmill for a pre-
set distance of air travel. See Section 4.2.

Demagnetizing. Refers to the process of reducing the
magnetic pole strength of a generator rotor to the
appropriate value.

Detonator. A device which initiates an explosive
train in response to an electrical current.

Diode. A two-element electron tube used as the
rectifier of the doppler signal.

Diode Impedance. The resistance which, when in-
serted in series with a perfect diode, would make it
behave like an actual diode.

Directivity Pattern. A polar plot of the antenna gain
as a function of angle. See f2(0) in Section 2.8.

Doppler Effect. The shift in frequency produced by
relative motion between transmitter and receiver.

Doppler Frequency. The difference or shift frequency
produced by the doppler effect.

Driver. A term applied to the windmill or turbine
used as the prime mover of the generator.

Dud. A fuze which does not function.

433

434

GLOSSARY

Dumping. The discharge of the thyratron plate con-
denser (in a fuze having RC arming) with in-
sufficient current to fire the detonator. See Section
3.3.6.

E. Early function. Function or operation of the fuze
after arming and before reaching the target some-
times, particularly in the case of rockets, confined
to the first 5 seconds of flight.

Effective Critical Voltage. The critical voltage of
the thyratron in a fuze under operating conditions
but in the absence of target signal. See Section 3.3.2.

Electrical Arming. Completion of electrical circuits
necessary to produce arming in a bomb fuze. This
usually occurs slightly ahead of mechanical arming,
in bomb fuzes. In fuze with RC arming, electrical
arming occurs after mechanical arming.

. Emerson. Emerson Radio and Phonograph Corpora-
tion.

Feedback Network. The network in an amplifier
which returns a portion of the output signal to the
amplifier input for controlling the gain frequency
characteristic. See Section 3.2.3.

Field Test. A functional test of a fuze on a missile
in flight.

Final Test or Final Test Position. Laboratory test
or equipment for making the test on the completed
electronic subassembly of the fuze. See Section 7.9.

Firing Indicator. A laboratory device (usually com-
prising a neon lamp) to indicate the firing of the
thyratron.

Firing Voltage. The signal input to an amplifier re-
quired to fire the thyratron, usually measured in
millivolts. Cf. MvF.

FOMA. Function on mechanical arming. See Section
8.3.

Frequency, Audio or Doppler. See doppler frequency.

Frequency, Carrier. See carrier frequency.

Frequency, Generator. The frequency of the alter-
nating current produced by the generator.

Frequency, Microphonic. The frequency of the
microphonic disturbances.

Frequency, Rotational. The rotational frequency of
the wind-driven generator system.

Frequency Response Curve. A plot of amplifier gain
versus audio frequency, usually plotted on log-log
paper.

g. Unit of acceleration; viz., 32.2 feet per second per
second.

Gain, Antenna. See antenna gain.

Gain, Audio. See audio gain.

Gain, Flat. The gain of the audio amplifier with the
feedback network disconnected.

Gain, Nonregenerative. See gain, flat.

Gain-Frequency Characteristics. See frequency re-
sponse curve.

Gear Train. A speed-reducing train of gears inter-
posed between the windmill and mechanical arming
system.

Gimmick. A small variable capacitor composed of

twisted insulated wires, used to adjust the amplifier
feedback.

Green. The code term for 150 megacycles per second.

Grid Reaction. The variation of grid bias with
oscillator load. See Section 3.1.2.

HD. Heard dud. A dud in which the carrier signal of
the fuze is observed.

Head-on Doppler. A doppler frequency for head-on
approach to an airplane target.

Holding Bias. The excess of the applied thyratron
bias over the effective critical voltage.

Hum. The high-frequency audio signal output of the
amplifier occurring at generator frequency, and
usually due to modulation produced by a-c operation
of the filament.

Hum Injection. A portion of the filament voltage
injected into the amplifier to cancel the inherent hum.
See Section 3.2.3.

Impact Parameter. The perpendicular distance from
the missile trajectory to an airplane target.

Impeller. A term sometimes, but inaccurately, ap-
plied to a windmill or turbine prime mover generator.

Induction Field. That component of the transmitted
electromagnetic field which varies inversely as the
square of the distance from the transmitter. See
Section 2.10.

Jamming. A countermeasure to produce malfunction
of fuzes by radio methods.

L. Late. A fuze function lower than expected from
the normal distribution of function heights or (in
the case of an antiaircraft fuze) a function beyond
a statistically expected burst surface.

Load. Electrical load on an oscillator.

Load Resistor. A resistor used as an equivalent of
the antenna load. Cf. compensated resistor and Sec-
tion 7.2.

Longitudinal. Referring to a fuze system in which
the predominant radiating current flows along the
axis of the missile.

M. Middle function. A random function occurring more
than 5 seconds after launching.

M Wave. The actual audio signal produced by the
doppler effect of the target on the fuze. This term is
particularly applicable to the nonsinusoidal doppler
signal generated in approaching an airplane target.
See Sections 2.11 and 2.12.

Mechanical Arming. Removal of the mechanical
barrier in the explosive train. In the case of RC arm-
ing, it usually implies also the closing of the elec-
trical contacts to initiate the RC arming cycle.

Michigan Sensitivity. The theoretical height at
which a fuze would function when approaching
ground at optimum velocity with optimum orienta-
tion of the missile, usually approximately horizontal.

MRLG. (Mortar, Radio, Longitudinal, Generator.)
An early designation for the T-132 mortar fuze.

MROG. (Mortar, Radio, 0 for loop, Generator.) An
early designation for the T-172 mortar fuze.

Mutual Interference. Jamming of one fuze by an-

GLOSSARY

435

other in too close proximity. Cf. sympathetic func-
tion.

MvF. Cf. firing voltage. Generally used as inverse
measure of amplifier gain.

A7. Number. Usually referring to number of rounds
in a field test.

NBS. National Bureau of Standards.

NDRC. National Defense Research Committee.

Normal Critical Voltage. The thyratron critical
voltage in the absence of microphonics from the
oscillator.

Normalization. The process of demagnetizing a gen-
erator rotor to give correct supply voltages.

OD. Designation for the oscillator-diode type of fuze.

ODD. Ordnance Development Division of the National
Bureau of Standards.

OSRD. Office of Scientific Research and Development.

P. Proper function of the fuze on approach to the
target.

Peak Gain. The maximum gain of an amplifier.

Philco. Philco Radio and Television Corporation.

PkMvF. Millivolts to fire at the frequency of peak
gain. Cf. MvF.

Plate Reaction. The variation of the oscillator plate
current in response to a change of high-frequency
load on the oscillator.

Potato Masher. A term applied to the encasing can
for generator-powered bomb fuzes. See Figure 18,
Chapter 4.

POA. Puff on approach. Refers to spotting charge in-
dication of function on approach to target.

POD. Power-oscillating detector. A type of plate re-
action oscillator operating at relatively high power
level.

POW. Puff on water. Cf. POA.

Predicted Height. Height of function predicted on
the basis of laboratory measurements of overall fuze
sensitivity.

Propeller. A term used to indicate the externally
mounted windmill or driver for the generator.

Pulse Test. An overall test of a fuze assembly com-
plete except for detonator, indicating that all cir-
cuits are functioning.

Purge Pellet. A pellet of specially rapid burning
material to purge the combustion chamber of residual
propellant after the burning is essentially complete.
See Section 9.2 on afterburning.

Q. A figure of merit for a tuned circuit or a reactor.

Quasi Static. The component of the electromagnetic

field which varies inversely as the cube of the distance
from the transmitter. See Section 2.10.

R. Resistance.

Radiation Field. That component of the transmitted
electromagnetic field which varies inversely as the
distance from the transmitter. See Section 2.8.

Radiation Load. Radiation resistance. The com-
ponent of the antenna resistance representing radia-
tion losses.

Radiation Pattern. A polar plot of radiation field
strength versus angle, /(0). See Section 2.8.

Random Function. A fuze function after arming and
before the target is reached.

RC. Resistance-capacitance network or the time con-
stant of such a network used for time delay of elec-
trical arming in some fuze models. See Section 3.3.6.

Raytheon. Raytheon Manufacturing Company.

Red. Code designation for 130 megacycles per second.

Reflection Coefficient. Ratio of reflected to incident
field strength.

Regulation Circuit. A resistance-capacitance net-
work incorporated in the power supply to maintain
an output voltage or voltages essentially independent
of generator speed above a limited minimum speed.

Resistance Sensitivity. A measure of the differential
voltage produced by an oscillator under change of
load resistance. See Section 3.1.2.

RGD. Reaction grid detector. Cf. grid reaction.

Ring Type. A generator-powered fuze in which the
antenna consists of a ring surrounding the windmill.

Ripple Voltage. See hum voltage.

ROA. Radius of action. The maximum radius of a
hypothetical burst surface enclosing an airborne
target. See also the definition in Section 5.1.3.

ROB. Radio-operated bomb. A very early generic
designation of radio proximity fuzes for bombs.

ROR. Radio-operated rocket. Cf. ROB.

Rotor, Detonator. The moving portion of the me-
chanical barrier to an explosive train (in bomb
fuzes) .

Rotor, Generator. The rotating permanent magnet
in a generator.

RRLG. (Rocket, Radio, Longitudinal, Generator.) An
early designation for generator-powered proximity
fuzes for rockets.

Safe. The condition of a fuze which is not armed.

SD. Self destruction. Destruction of a fuze by opera-
tion of a device within the fuze at a predetermined
time or distance after launching, presumably after
the missile has passed the target. See Sections 4.3.1
and 3.3.8.

Sensitivity. See resistance sensitivity and Section
3.1.2.

Sensitivity Pattern. A hypothetical surface sur-
rounding a missile representing the locus of target
positions for functions.

Serpentine. A type of single coil generator winding.
See Section 3.4.5.

Setback. A term referring to reaction on a fuze or
missile caused by acceleration of the projectile.

Signal Simulator. A laboratory device to simulate
the signal produced by interaction of the fuze and
target. See Section 2.12.

Spikes. Short duration pulses, usually originating in
triode microphonics.

Squegging. The periodic instability of a high-
frequency oscillator. See Section 3.1.5.

Squib. Colloquial for electric detonator.

Surge Current. The peak value of a thyratron plate
current surge.

Sylvania. Sylvania Electric Products Corporation.

436

GLOSSARY

Sympathetic Function. The functioning of a fuze
on a spurious target produced by the explosion of
another missile.

T. Refers to target function in a field test.

Target Factor. Reflection coefficient of a ground tar-
get multiplied by 100.

Target Function. The proper function of a fuze
upon approach to the intended target.

Turbine. An air-driven turbine used in some models
of generator-power fuzes.

Vane. See windmill.

VT. The commonly accepted designation for proximity
fuze. It is generally understood that the letters stand

for “variable time” but they also imply “vacuum
tube.”

Westinghouse. Westinghouse Electric Corporation.

White. A code designation for 110 megacycles per
second.

Windmill. An externally mounted air-driven prime
mover.

Wurlitzer. The Rudolph Wurlitzer Corporation.

Yellow. A code designation for 140 megacycles per
second.

Zenith. Zenith Radio Corporation.

Zero Length Launcher. A launcher for rockets in
which the rocket is supported over a negligible por-
tion of its burning period.

BIBLIOGRAPHY

Numbers such as Div. 4-100-1M indicate that the document listed has been microfilmed and that its title
appears in the microfilm index printed in a separate volume. For access to the index volume and to the
microfilm, consult the Army or Navy agency listed on the reverse of the half-title page.

Chapter 1

1. ‘‘Notes on Conference of 12 August 1940 between
representatives of NDRC and BuOrd,” by Com-
mander C. Hoover, Aug. 17, 1940. Div. 4-100-MI

2. “Development of Anti-Aircraft Fuzes for Rockets,”
Initiation of Project, Ordnance Committee Min-
utes No. 18178, Apr. 24, 1942.

3. “Initiation of Radio Rocket Fuze Project,” Ord-
nance Committee Minutes No. 18364, June 5, 1942.

4. “Initiation of Development Project for T40, T43
Fuzes,” Ordnance Committee Minutes No. 19939,
Mar. 17, 1943.

5. “Initiation of Development Project for T50, T51,
T52 Fuzes,” Ordnance Committee Minutes No.
21117, July 17, 1943.

6. “Initiation of Development Project for T6 Fuzes,”
Ordnance Committee Minutes No. 21681, Sept. 8,
1943.

7. “Initiation of Development Project for T30 Fuzes,”
Ordnance Committee Minutes No. 25243, Sept. 28,
1943.

8. “Transfer of T82 Fuzes to Army Development,”
Ordnance Committee Minutes No. 26818, Mar. 1,
1943.

9. “Initiation of Development Project for T32, T2005
Fuzes,” Ordnance Committee Minutes No. 27280,
Apr. 12, 1945.

10. “Initiation of Development for T132, T171, T172
Fuzes,” Ordnance Committee Minutes No. 27427,
Apr. 26, 1945.

11. The Optimum Point of Burst for a 500-lb GP
Bomb Equipped with a Proximity Fuze, by
Marston Morse, William R. Transue, Roy Kuebler,
TDBS Report 7, Office of the Chief of Ordnance,
Apr. 22, 1943.

12. The Dependence of Optimum Height of Burst of
Shells and Bombs upon Angle of Fall, Safety
Angle, etc., by Marston Morse, William R. Tran-
sue, TDBS Report 41, Office of the Chief of Ord-
nance, Sept. 2, 1944.

13. Optimum Height of Burst of Fragmentation
Bombs and Effect with VT Fuzes, by Marston
Morse, William R. Transue, and M. H. Heins,
TDBS Report 58, Office of the Chief of Ordnance,
Apr. 3, 1945.

14. Probable Advantages of VT Fuzes on 81-mm HE
Mortar Shell M56 and MU3A1, Marston Morse,
William R. Transue, and M. H. Heins, TDBS Re-
port 60, Office of the Chief of Ordnance, Mar. 30,
1945.

15. A Comparison of Damage Effect of Ground Bursts
of 20-lb Bomb with an Air Burst of the 260-lb

Bomb and of the 500-lb Bomb against Planes and
Revetments, TDBS Report 61, Office of the Chief
of Ordnance, Apr. 24, 1945.

16. Second Interim Report on Fuze, Bomb, T50, Re-
port of the Army Air Force Proving Ground Com-
mand on Project 4012C4712.82, Apr. 12, 1945.

17. Supplemental Tests on Aircraft Rockets for Anti-
Personnel Effects, Report of the Army Air Forces
Proving Ground Command on Project 4514C471.94,
Sept. 4, 1945.

18. Final Report on Air-to-Air Firing of Mk 171
Mod 0 Fuzes on 3.U" and 5.0" AR, NOTS Project
104 AFS, Aug. 5, 1945.

19. Probability that a 1+.5" Rocket Fired from Astern

Will Destroy a Twin-Engine Bomber (JU-88) as a
Function of Point of Burst, AMP Report 21. 1R,
Statistical Research Group, Applied Mathematics
Panel, July 1944. Div. 4-412.3-MI

20. Optimum Burst Surface for U.5" Airborne Rocket
Fired from Astern at Twin-Engine Bomber (JU-
88), AMP Report 21.2R, Statistical Research
Group, Applied Mathematics Panel, July 1944.

Div. 4-412.3-M2

21. Effectiveness of a U.5" Airborne Rocket with T5
Fuze When Fired at Twin-Engine Bomber from
Astern, AMP Report 21. 3R, Statistical Research
Group, Applied Mathematics Panel, July 1944.

Div. 4-412.3-M3

22. Probability' of Damage Computations Pertinent to
Design of Fuze for 5" AR and 5" HV AR, Milton
Friedman, Informal Study (AMP Study 21, SRG
396), Applied Mathematics Panel, January 1945.

Div. 4-412.3-M5

23. Airburst for Blast Bombs, E. Bright Wilson, Jr.,

OSRD 4943, OEMsr-260 and OEMsr-569, Service
Projects OD-03, NO-224, et al., Division 2. Report
A-322, Princeton University, WHOI, et al., April
1945. Div. 4-242. 12-M4

24. Evaluation of Airburst Bombs for Clearance of
Mine Field, Robert D. Huntoon, OSRD 4100, Serv-
ice Project OD-27, Report A-291, September 1944.

Div. 4-242.12-MI

25. Effect of Height of Detonation of Bombs on the
Blast Pressures and Impulses of Surrounding
Buildings, in Richmond Park 1/7 Square Model
Town Tests, Road Research Laboratory, Depart-
ment of Scientific and Industrial Research Minis-
try of Supply, Note MOS/434/RJ.EK, March 1945.

26. Trials with an M6U 500-lb Bomb, Nose Initiated,
Fuze T50 against Close Support Targets , B. L.
Welch, Appendix to Proceeding Q-2881, Ordnance
Board, Dec. 13, 1944.

27. Note on Airbursts of U, 000-lb HC Bomb with T51

437

438

BIBLIOGRAPHY

Fuze , F. H. East, Technical Note ARM343, Royal
Aircraft Establishment, April 1946.

28. Optimum Height of Setting fo r T50 Fuze on Blast
Bombs, A 1C LC 500-lb Mark 2 Charged Dyed
Methyl Scelicyliate and Dropped onto Jungle, San
Jose, Project Report 69, Chemical Warfare Serv-
ice, June 22, 1945.

29. Multiple Bomb Assessment of Blast Bomb A/C
LC 500-lb Mark 2 Fitted with T51 Fuze and
Charged HT when Dropped from High Altitudes
into Jungle Terrain, San Jose, Project Report 73,
Chemical Warfare Service, July 28, 1945.

30. Interim Report, February 15 to March 7, 1945,
A. V. Astin to Dr. Alexander Ellett.

31. Fire Bombs Tried at Eglin Field with VT Fuzes,
T. N. White, Report OD-2-255M, NBS, Ordnance
Development Division, July 13, 1945.

Div. 4-242.13-M2

32. Operational Uses of Bomb and Rocket VT Fuzes
by U. S. Army and Navy in World War II, Walter
G. Finch, Report OD-Army-4, NBS, Ordnance De-
velopment Division, Oct. 15, 1945. Div. 4-221-M2

33. Letter to R. C. Tolman. Subject: “The Use in
Proximity Fuzes for Rockets of the Various Elec-
tronic Components of Small Size for Use in the
Proximity Fuze for Antiaircraft Projectiles,”
W. S. Parsons, May 21, 1942. Div. 4-211.1-MI

34. Evaluation of Air Burst Bombs for Clearance of
Mine Fields, E. F. Horton, Jr., Final Report on
Experimental Investigation, OD-1-599, NBS, Ord-
nance Development Division, Dec. 23, 1944.

Div. 4-242. 12-M3

Chapter 2

ARMOR AND ORDNANCE REPORTS OF NDRC

1. Radio Controlled Antiaircraft Proximity Fuze;
The Reflection of Radio Waves from Airplanes,
Robert D. Huntoon, Service Project OD-27, Prog-
ress Report A-19, Nov. 10, 1941. Div. 4-211-MI

2. Radio Proximity Fuzes for Bombs and Rockets as

of May 28, 1942, Harry M. Diamond, Service Proj-
ects OD-27, OD-33, and CWS-19, Progress Report
A-64, June 15, 1942. Div. 4-211.1-M2

3. A Device for the Measurement of the Absolute

Sensitivity of an End-Fed Axially -Excited Radio
Proximity Fuze, William L. Kraushaar, and Rob-
ert D. Huntoon, Service Project OD-27 and OD-26,
Report A-143, Feb. 13, 1943. Div. 4-625-Ml

4. Radio Proximity Fuze for Plane-to-Plane Rocket
Application, Harry M. Diamond, W. S. Hinman,
Jr., Robert D. Huntoon, Cledo Brunetti, and Ches-
ter H. Page, Service Projects OD-27 and OD-26,
Report A-144, Feb. 12, 1943. Div. 4-211. 1-M3

5. VT Fuzes for Rockets and Bombs, Training Lec-
tures, Robert D. Huntoon, Chester H. Page, B. J.
Miller, Jacob Rabinow, and Harry M. Diamond,
OSRD 5326, Service Projects OD-27, NO-77B, and
NO-77R, Report A-334, January 1945.

Div. 4-200-MI

6. Radiation Properties of BRLG, Robert D. Huntoon,
Service Project OD-27, Report 43-R, July 28, 1943.

Div. 4-243. 11-Ml

7. Design of Special Targets, Robert D. Huntoon,
Service Project OD-27, Report 44-R, May 12, 1943.

Div. 4-618-M2

8. Selection of Optimum Frequencies for BRLG

Vehicles, Robert D. Huntoon, Service Project
OD-27, Report 52-R, August 1943. Revised: Apr.
17, 1944. Div. 4-243. 11-M2

REPORTS OF ORDNANCE DEVELOPMENT DIVI-
SION OF NATIONAL BUREAU OF STANDARDS

9. Basic Theory of the Radio Proximity Fuze, Philip

R. Karr, NBS, Ordnance Development Division,
May 25, 1945. Div. 4-211-M2

10. Afterburning from Rocket Motors and Malfunc-
tioning of VT Fuzes (Summary Report), H. F.
Stimson, Report OD-1-896, NBS, Ordnance De-
velopment Division, Oct. 15, 1945.

Div. 4-411. 11-M6

11. Theoretical Estimates of the Radiation Resistance
of the BRTG Propeller Antenna Model, J. G. Hoff-
man and David Feldman, Report OD-2-30, NBS,
Ordnance Development Division, Apr. 24, 1944.

Div. 4-243. 21-M2

12. Prediction of Heights of Function (Supplement
to Report OD-3-89), Bertrand J. Miller and
M. Schulkin, Report OD-BE-22R, NBS, Ordnance
Development Division, Aug. 11, 1944.

Div. 4-241-M3

13. Electrical Interaction of T-50 Fuzes (Part II),
Bertrand J. Miller, Report OD-BE-42R, NBS, Ord-
nance Development Division, Sept. 29, 1944.

Div. 4-245-M4

14. Expected Radius of Action for the T-30 , Bertrand
J. Miller and Franklin M. Fletcher, Report OD-
BE-53R (and Addendum), NBS, Ordnance De-
velopment Division, Nov. 11, 1944.

Div. 4-241.1-MI

15. Estimates of Radius of Action of T-30 from Steady

State Computations, R. F. Morrison, Jr., Thomas
M. Marion, and Franklin M. Fletcher, Report OD-
BE-56R, NBS, Ordnance Development Division,
Dec. 4, 1944. Div. 4-241.1-M3

16. Construction of Apparatus for Measuring Reflec-

tion Coefficient, Otto E. Spokas, Report OD-BE-
77R, NBS, Ordnance Development Division, Apr.
23, 1945. Div. 4-624-M3

17. Measurement of the Reflection Coefficient of the
Water Bombing Range at Aberdeen Proving
Ground, Otto E. Spokas, Report OD-7-201R, NBS,
Ordnance Development Division, May 1, 1945.

Div. 4-624-M4

18. Striking Angles and Velocities for Level Flight
Bombing, Allen T. Foster, Report OD-7-87R, NBS,
Ordnance Development Division, Mar. 20, 1945.

Div. 4-311.3-M2

19. Impact Angles and Striking Velocities for Dive

BIBLIOGRAPHY

439

Bombing, F. L. Celauro, Report OD-7-88R, NBS,
Ordnance Development Division, Mar. 22, 1945.

Div. 4-242.14-MI

20. Striking Angles and Velocities for Level Flight

Bombing with the M-65 Bomb, Allen T. Foster,
Report OD-2-223R (Supplement to OD-7-87R),
NBS, Ordnance Development Division, June 5,
1945. Div. 4-311.3-M3

21. Striking Angles and Velocities for Level Flight
Bombing with M-57, Allen T. Foster, Report OD-2-
257R (Supplement to OD-7-87R), NBS, Ordnance
Development Division, July 18, 1945.

Div. 4-311.3-M4

22. Effect of Ground Reflection on BRLG Perform-

ance, Charles J. Apolenis and Robert D. Huntoon,
Report OD-3-19, NBS, Ordnance Development
Division, Nov. 2, 1943. Div. 4-243.21-MI

23. Induction Field Sensitivity , Chester H. Page, Re-

port OD-3-30, NBS, Ordnance Development Divi-
sion, Nov. 16, 1943. Div. 4-233-MI

24. Induction Field of Finite Antennas, Chester H.

Page, Report OD-3-33, NBS, Ordnance Develop-
ment Division, Nov. 19, 1943. Div. 4-233-M2

25. Experimental Measurement of the Effect of an

Imperfect Reflector on the Induction Field Sensi-
tivity of a Radio-Proximity Fuze, Otto E. Spokas,
Charles C. Gordon, and Robert D. Huntoon, Re-
port OD-3-36, NBS, Ordnance Development Divi-
sion, Nov. 25, 1943. Div. 4-624-MI

26. Computation of Heights of Function, Including

Induction and Quasi-Static Field Effects, Bertrand
J. Miller and Philip R. Karr, Report OD-3-89,
NBS, Ordnance Development Division, Jan. 29,
1944. Div. 4-241-M2

27. Measurement of the Reflection Coefficient of the
New Bombing Range at Aberdeen Proving Ground,
Otto E. Spokas, Report OD-3-90, NBS, Ordnance
Development Division, Jan. 29, 1944.

Div. 4-624-M2

28. Radiation Resistance of [the M-9] Rocket, Otto
E. Spokas, Charles C. Gordon, and Robert D.
Huntoon, Report OD-3-105, NBS, Ordnance De-
velopment Division, Mar. 2, 1944.

Div. 4-243. 22-MI

29. Microphonic Stability of Oscillator-Diode Type of

Fuze Circuit, Robert D. Huntoon, Report OD-3-
117, NBS, Ordnance Development Division, Mar.
22, 1944. Div. 4-238.31-Ml

30. Dummy Antennas, Robert D. Huntoon, Report

OD-3-133, NBS, Ordnance Development Division,
Apr. 20, 1944. Div. 4-233-M3

31. Tuning Compromise for BRLG Units, Philip R.

Karr and Otto E. Spokas, Report OD-3-139, NBS.
Ordnance Development Division, May 2, 1944.
Revised: June 3, 1944. Div. 4-233. 1-M5

32. Compensated Resistors for Tuning and Loading

Standards, E. Eisner and Paul T. Hawes, Report
OD-3-154, NBS, Ordnance Development Division,
May 24, 1944. Div. 4-236-M4

33. Antenna Rings for BRLG, Philip Krupen, Report

OD-3-162, NBS, Ordnance Development Division,
June 15, 1944. Div. 4-233-M4

34. RGD Field Simulator, Philip Krupen, Report OD-

3-163, NBS, Ordnance Development Division, June
17, 1944. Div. 4-238. 32-M7

35. Pole Tests on Various Vehicles at Blossom Point,
James H. Barnard, Glenn L. Scillian, and Ralph
Stair, Report OD-3-174, NBS, Ordnance Develop-
ment Division, Aug. 16, 1944. Div. 4-243.4-M2

36. Radiation Patterns and Electrical Balance of

BRTG, Glenn L. Scillian and Ralph Stair, Report
OD-3-177, NBS, Ordnance Development Division,
Aug. 31, 1944. Div. 4-243.11-M5

37. Radiation Resistance of Zenith BRTG-Z Units,

Glenn L. Scillian and Ralph Stair, Report OD-3-
178, NBS, Ordnance Development Division, Sept.
13, 1944. Div. 4-243. 21-M4

38. Resonant Loading of BRTG Units by Test Boxes,
Ralph Stair, Glenn L. Scillian, and Leonard C.
Pochop, Report OD-3-196, NBS, Ordnance De-
velopment Division, Nov. 13, 1944.

Div. 4-233.1-M7

39. Transparent Charts for Prediction of Function
Height, Philip R. Karr, Chris Gregory, R. B.
Schwartz, and M. L. Scott, Report OD-3-257, NBS,
Ordnance Development Division, June 6, 1945.

Div. 4-241-M7

40. Low Frequency Operation of Bomb Fuzes, R. B.
Schwartz, Report OD-3-258, NBS, Ordnance De-
velopment Division, June 7, 1945.

Div. 4-243. 11-Mil

41. Computation of Burst Heights of Longitudinally -
Excited Bomb Fuzes, R. B. Schwartz, Report OD-
3-281, NBS, Ordnance Development Division, Aug.

7, 1945. Div. 4-241-M8

42. Early Functions of the MC-382 Radio-Operated
Plane-to-Plane Rocket Fuze, Bertrand J. Miller
and Robert D. Huntoon, Progress Report OD-3-
AB2, NBS, Ordnance Development Division, June

8, 1943. Div. 4-222. 128-M12

MEMORANDA OF ORDNANCE DEVELOPMENT DI-
VISION OF NATIONAL BUREAU OF STANDARDS

43. Amplifier Shaping and After-Burning Pulses,

Memorandum to Robert D. Huntoon from Ber-
trand J. Miller, NBS, Ordnance Development Divi-
sion, Mar. 4, 1943. Div. 4-238.212-MI

44. After-Burning and Amplifier Shaping, Memoran-

dum to W. S. Hinman, Jr., from Robert D. Hun-
toon, NBS, Ordnance Development Division, Mar.
5, 1943. Div. 4-238. 212-M2

45. After-Burning, Memorandum to W. S. Hinman,
Jr., from Robert D. Huntoon, NBS, Ordnance De-
velopment Division, Mar. 18, 1943.

Div. 4-411. 11-MI

46. Pole Tests on British Two-Ton Vehicle, Memoran-
dum to A. V. Astin from Ralph Stair and James

440

BIBLIOGRAPHY

H. Barnard, Memorandum OD-3-33M, NBS, Ord-
nance Development Division, Aug. 24, 1944.

Div. 4-243.4-M3

47. Radiation Patterns on Zenith and Westinghouse,
BRTG, Memorandum to A. V. Astin from Ralph
Stair Memorandum OD-3-34M, NBS, Ordnance
Development Division, Aug. 24, 1944.

Div. 4-243. 11-M3

48. Computation of Expected Radius of Action, Memo-
randum to Harry M. Diamond from Chester H.
Page, Memorandum OD-3-53M, NBS, Ordnance
Development Division, Nov. 6, 1944.

Div. 4-241.1-M2

49. Radiation Resistance of BRLG Vehicles, Memo-
randum to Harry M. Diamond from Robert D.
Huntoon, Memorandum OD-BE-2M, NBS, Ord-
nance Development Division, June 20, 1944.

Div. 4-243. 21-M3

50. Electrical Properties of British 4, 000-lb Bomb,
Memorandum to Alexander Ellett from Harry M.
Diamond, Memorandum OD-BE-42M, NBS, Ord-
nance Development Division, Aug. 26, 1944.

Div. 4-243.1 1-M4

51. Mutual Interaction in BRLG Units Dropped in

Close Spaced Train, Memorandum to Harry M.
Diamond from Bertrand J. Miller, Memorandum
OD-BE-44M, NBS, Ordnance Development Divi-
sion, Sept. 11, 1944. Div. 4-245-M3

52. Radiation Properties of British U, 000-lb Bomb,

Memorandum to Harry M. Diamond from Frank-
lin M. Fletcher and Otto E. Spokas, Memorandum
OD-BE-47M, NBS, Ordnance Development Divi-
sion, Sept. 9, 1944. Div. 4-243. 11-M6

53. Interaction Factors for BRLG Units, Memoran-
dum to Harry M. Diamond from Franklin M.
Fletcher, Memorandum OD-BE-48M, NBS, Ord-
nance Development Division, Sept. 11, 1944.

Div. 4-245-M2

54. Radiation Properties of HVAR 5" Rocket, Memo-

randum to Harry M. Diamond from Otto E.
Spokas and R. F. Morrison, Jr., Memorandum OD-
BE-50M, NBS, Ordnance Development Division,
Sept. 13, 1944. Div. 4-243.22-M2

55. Radiation Properties of Depth Bombs, Memoran-

dum to Harry M. Diamond from Otto E. Spokas
and Franklin M. Fletcher, Memorandum OD-BE-
53M, NBS, Ordnance Development Division, Sept.
15, 1944. Div. 4-243. 11-M7

56. Radiation Properties of MU3 and M56, Memoran-

dum to Harry M. Diamond from Otto E. Spokas
and Franklin M. Fletcher, Memorandum OD-BE-
54M, NBS, Ordnance Development Division, Sept.
18, 1944. Div. 4-243. 13-MI

57. Additional Measurements on Radiation Properties
of the British U, 000-lb Bomb (Supplement to OD-
BE-47M), Memorandum to Harry M. Diamond
from Otto E. Spokas and Franklin M. Fletcher,
Memorandum OD-BE-56M, NBS, Ordnance De-
velopment Division, Sept. 19, 1944.

Div. 4-243. 11-M8 I

58. Radiation Properties of 1,000 and 2, 000-lb GP

Bombs, Memorandum to Harry M. Diamond from
Otto E. Spokas and Franklin M. Fletcher, Memo
randum OD-BE-59M, NBS, Ordnance Development
Division, Sept. 27, 1944. Div. 4-243.11-M9

59. Radiation Properties of the 5-inch Mattress and

the 155 mm Mortar Projectile, Memorandum tc
Harry M. Diamond from Otto E. Spokas ana
Franklin M. Fletcher, Memorandum OD-BE-63M,
NBS, Ordnance Development Division, Sept. 30,
1944. Div. 4-243. 13-M2

60. Radiation Properties of Vehicles M30, M6U, and
M81, Memorandum to Harry M. Diamond from
Franklin M. Fletcher and Otto E. Spokas, Memo-
randum OD-BE-66M, NBS, Ordnance Develop-
ment Division, Oct. 5, 1944. Div. 4-243. 11-M10

61. Calculations Concerning Radius of Action in
Plane-to-Plane Application, Memorandum to
Harry M. Diamond from Bertrand J. Miller,
Memorandum OD-BE-82M, Nov. 14, 1944.

Div. 4-412.4-M4

62. Radiation Properties of Gas Tanks, Memorandum
to Harry M. Diamond from Bertrand J. Miller,
Preliminary Memorandum OD-BE-89M, NBS,
Ordnance Development Division, Nov. 27, 1944.

Div. 4-243.3-MI

63. Radiation Properties of Various Rockets, Memo-
randum to Harry M. Diamond from Bertrand J.
Miller, Memorandum OD-BE-92M, NBS, Ord-
nance Development Division, Dec. 12, 1944.

Div. 4-243. 22-M3

64. Radiation Resistance of the M56 Mortar, the MU3
Mortar with an M56 Tail, the AN -M U1 Fragmen-
tation Bomb and the 155 mm Chemical Mortar
Projectile When Used with an MRLG Type Unit,
Memorandum to Harry M. Diamond from Otto E.
Spokas, Memorandum OD-BE/-98M, NBS, Ord-
nance Development Division, Dec. 19, 1944.

Div. 4-243. 23-MI

65. The Effect of Various Antenna Rings on the Radi-

ation Resistance of the M56 Mortar and the MU3
Mortar with the M56 Tail, Memorandum to Harry
M. Diamond from Otto E. Spokas, Memorandum
OD-BE-127M, NBS, Ordnance Development Divi-
sion, Apr. 2, 1945. Div. 4-243.23-M2

66. Additional Radiation Resistance Data on the

HVAR, AR3.5 and AR5 Rockets, Memorandum to
Harry M. Diamond from Otto E. Spokas, Memo-
randum OD-7-202M, NBS, Ordnance Development
Division, May 1, 1945. Div. 4-243. 22-M4

67. Radiation Resistance Presented to the Type T-2005
Unit, Memorandum to Harry M. Diamond from
Otto E. Spokas, Memorandum OD-7-205M, NBS,
Ordnance Development Division, June 25, 1945.

Div. 4-243. 22-M5

68. Radiation Patterns of the AR and Hb.5 Rocket,
Memorandum to Harry M. Diamond from Otto E.
Spokas, Memorandum OD-7-212M, NBS, Ordnance
Development Division, July 21, 1945.

Div. 4-243. 12-M2

BIBLIOGRAPHY

69. Radiation Resistance of the M-56 Mortar Shell

with 2 " Tail Extension, Memorandum to Harry M.
Diamond from Otto E. Spokas, Memorandum OD-
7-213M, NBS, Ordnance Development Division,
Aug. 28, 1945. Div. 4-243.23-M3

70. Revision and Extension of OD-OAG-20 (Striking
Velocity, Striking Angle, Vei'tical Component of
Striking Velocity vs Altitude), Memorandum to
Recipients of OD-OAG-20 from D. Fisher, OD-
OAG-41, NBS, Ordnance Development Division,
Sept 11, 1944.

71. Tables of Doppler Frequency vs Altitude of Re-
lease at 200 Miles Per Hour for Carrier Fre-
quencies, D. Fisher, Report OD-OAG-42, NBS,
Ordnance Development Division, Sept. 19, 1944.

Div. 4-412.4-M2

72. Summary of Ballistic Data for the Mk-7, CIT
Rocket, Memorandum to Harry M. Diamond from
F. A. Ransom, Memorandum OD-OAG-45, NBS,
Ordnance Development Division, Sept. 28, 1944.

Div. 4-412. 4-M3

73. Navy Rocket Trajectory Analysis, Memorandum

to T. N. White from A. L. Leiner (in collabora-
tion with D. C. Friedman), Memorandum OD-2-
203, NBS, Ordnance Development Division, May
5, 1945. Div. 4-412.1-M8

74. Table of Bomb Velocity vs Air Travel, Allen T.
Foster, Report OD-2-252M, NBS, Ordnance De-
velopment Division, July 5, 1945. Div. 4-242.14-M2

75. Striking Angles and Vertical Components of Strik-
ing Velocities of Rockets Fired from an Airplane
in Dive. Memorandum to T. N. White from F. L.
Celauro, Memorandum OD-2-261M, NBS, Ord-
nance Development Division, July 25, 1945.

Div. 4-412. 4-M6

REPORTS OF CONTRACTORS OF DIVISION U
OF NDRC

76. Research and Development Conducted by Philco

Corporation on PU-772 Radio Proximity Fuze for
Large Bombs (Final Report), R. A. Bell, OEMsr-
866, Symbol 2164, Philco Radio and Television
Corporation, June 15, 1943. Div. 4-211. 1-M4

77. Considerations of the Problem of Adapting the

Radio Proximity Fuze to the M-56 Mortar Pro-
jectile, Alfred S. Khouri, University of Florida,
Oct. 30, 1943. Div. 4-211.23-MI

78. A Study of the Possibility of Making Both the
Loop and Longitudinal Type Fuzes from the Basic
University of Florida MROG Unit, Alfred S.
Khouri, University of Florida, Apr. 3, 1945.

Div. 4-211. 23-M5

79. Modifications of the MROG to Reduce the Loop
Area and Prominence of the Loop, Alfred S.
Khouri, University of Florida, Mar. 29, 1945.

80. Mortimer Loop Radio Proximity Fuze Report,
University of Florida, Apr. 22, 1944.

Div. 4-211. 23-M2

81. Interaction of Loop Antenna and Neighboring
Conductors with Special Reference to the MROG

%Fuze, R. C. Williamson, Report WRL-UF-3jJJni-
versity of Florida, Aug. 10, 1944. Div. 4-2K-M5

441

82. A Possible Method of Reducing the Undesired

Parasitic Radiation from a Vehicle Excited Trans-
versely, C. Albert Moreno, University of Florida,

Nov. 1, 1943. Div. 4-243.4-MI

83. Performance of the Basic MROG Design Adapted
to End-Fed Longitudinal Excitation, Alfred S.
Khouri, University of Florida, Apr. 12, 1945.

84. Final Technical Report on Generator-Powered

Proximity Fuzes for Bombs, K. D. Smith and
A. L. Stillwell, OEMsr-905, Bell Telephone Labo-
ratories, May 30, 1944. Div. 4-211. 21-M5

85. Generator Report, R. N. Harmon, Westinghouse

Electric and Manufacturing Company, Apr. 8,
1943. Div. 4-232.2-M2

86. Development of a Ground Approach Proximity

Bomb Nose Fuze, BRTG, T. M. Bloomer, OEMsr-
343 and OEMsr-1106, Termination Report CFE-
760, Westinghouse Electric and Manufacturing
Company, Apr. 28, 1945. Div. 4-211.21-M11

87. Proximity Fuze, Bomb, Nose, Ground Approach,

Type VT, T-82, T. M. Bloomer, OEMsr-343 and
OEMsr-1106, Termination Report CFE-759, West-
inghouse Electric and Manufacturing Company,
Apr. 28, 1945. Div. 4-222.113-M2

88. Proximity Fuze — Hornet, John R. Boykin, OEMsr-

343, Termination Report CFE-762, Westinghouse
Electric and Manufacturing Company, Apr. 28,
1945. Div. 4-211.1-M6

89. Proximity Fuze [ for the ] Plane-to-Plane Rocket,
Type POD, John R. Boykin, OEMsr-343, Termina-
tion Report CFE-761, Westinghouse Electric and
Manufacturing Company, Apr. 28, 1945.

Div. 4-211.1-M5

90. BRLG Proximity Fuzes (Final Report), F. H.
Osborne, OEMsr-1161 and OEMsr-1163, Rudolph
Wurlitzer Company, Mar. 15, 1945.

Div. 4-211. 22-MI

91. Generator-Powered Radio Proximity Fuze for
Bombs — Transverse Antenna Type, Earl J. Diehl,
OSRD 5111, OEMsr-980 and OEMsr-1133, Service
Projects OD-27 and NO-77B, Final Report A-326,
Zenith Radio Corporation, Mar. 30, 1945.

Div. 4-211.21-M10

REPORTS OF OTHER DIVISIONS OF NDRC

92. Repeater Jamming of Radio Proximity Fuzes, Rus-
sell Yost, Jr., and Walter E. Tolies, OEMsr-1305,
Service Projects SC-98.07 and NA-109, Division 15
Report 1305-26, Jan. 27, 1946. Div. 4-246-MI

BRITISH REPORTS

93. Reflections from Bodies, N. F. Mott, Apr. 24, 1941.

UNCLASSIFIED TECHNICAL PUBLICATIONS

94. “Circuit Relations in Radiating Systems and Ap-
plications to Antenna Problems,” P. S. Carter,
Proceedings of the Institute of Radio Engineers,
Vol. 20, No. 6, June 1932, pp. 1004-1041.

95. “The Reciprocal Energy Theorem,” J. R. Carson,
Bell System Technical Journal, Vol. 9, April 1930,
pp. 325-331.

442

BIBLIOGRAPHY

Chapter 3

ARMOR AND ORDNANCE REPORTS OF NDRC

1. Radio Controlled Antiaircraft Proximity Fuze:

The Reflection of Radio Waves from Airplanes ,
Robert D. Huntoon; based on cooperative work by
Harry M. Diamond, W. S. Hinman, Jr., Robert D.
Huntoon, Cledo Brunetti, and Chester H. Page,
Service Project OD-27, Progress Report A-19,
Nov. 10, 1941. Div. 4-211-MI

2. The Performance of Small Dry Batteries When

Subjected to Low Temperatures and the Effect of
Heating the Batteries Internally by Alternating
Current Supplied to the Battery Terminals, John
P. Schrodt, D. Norman Craig, and George W.
Vinal, Service Project OD-27, Progress Report
A-30, Jan. 20, 1942. Div. 4-232.1-MI

3. The National Bureau of Standards Battery for
Low Temperature Operation, John P. Schrodt,
D. Norman Craig, and George W. Vinal, Service
Project OD-27, Progress Report A-49, May 2, 1942.

Div. 4-232.1-M2

4. The Possibility of a Generator Power Supply for

Proximity Fuzes, Allen S. Clarke, Service Proj-
ects OD-27 and OD-33, Progress Report A-62M,
Dec. 15, 1942. Div. 4-232.2-MI

5. Radio Proximity Fuzes for Bombs and Rockets as

of May 28, 19U2, Harry M. Diamond, Service Proj-
ects OD-27, OD-33, and CWS-19, Progress Report
A-64, June 12, 1942. Div. 4-211.1-M2

6. Firing of Squibs by Condenser Discharge — Energy
Losses in Thyratrons, Evert G. Bennett and Rich-
ard K. Cook, Service Projects OD-27 and OD-33,
Progress Report A-65, June 25, 1942.

Div. 4-231. 1-MI

7. Circuit Design of the Ultra-High Frequency Unit

for the Radio Proximity Fuze, Chester H. Page,
Service Projects OD-27 and OD-33, Progress Re-
port A-80, Aug. 11, 1942. Div. 4-211. 2-MI

8. Characteristics of Small Thyratrons for Use in

Proximity Fuzes, Mahlon F. Peck, Service Projects
OD-27 and OD-33, Progress Report A-112, Nov.
10, 1942. Div. 4-231.1-M4

9. Analysis of the Feedback Amplifier for MC-382

Fuze, Robert D. Huntoon, William L. Kraushaar,
and Herbert D. Cook, Progress Report A-122,
Dec. 7, 1942. Div. 4-238.222-MI

10. A Device for the Measurement of the Absolute
Sensitivity of an End-Fed Axially -Ex cited Radio
Proximity Fuze, William L. Kraushaar and Rob-
ert D. Huntoon, Service Projects OD-27 and
OD-26, Report A-143, Feb. 11, 1943.

Div. 4-625-MI

11. Radio Proximity Fuze for Plane-to-Plane Rocket
Application, Harry M. Diamond, W. S. Hinman,
Jr., Robert D. Huntoon, Cledo Brunetti, and Ches-
ter H. Page, Service Projects OD-27 and OD-26,
Report A-144, Feb. 12, 1943. Div. 4-211. 1-M3

12. Generator Powered Radio Proximity Fuze for

Bombs; Transverse Antenna Type, Earl J. Diehl,
OSRD 5111, OEMsr-980 and OEMsr-1133, Service
Projects OD-27 and NO-77B, Final Report A-326,
Zenith Radio Corporation, Mar. 30, 1945.

Div. 4-211. 21-M10

13. Pilot Production of T-50 Fuzes, Allen S. Clarke
and C. N. Julian, OSRD 5351, OEMsr-1227, Serv-
ice Projects OD-27, NO-77B, and NO-77R, Report
A-335, Bowen and Company, Inc., Apr. 12, 1945.

Div. 4-222. 111-M3

14. A Radio Proximity Fuze: Type MROG, OSRD

5412, OEMsr-949, Service Project OD-27, Report
A-338, War Research Laboratory, University of
Florida, April 1945. Div. 4-211.23-M3

15. Specification for Rectifier Bridge Assembly RA-1,

NBS, Ordnance Development Division, July 5,
1944. Div. 4-235-M6

16. Specification of Generator G-l, NBS, Ordnance
Development Division, Dec. 9, 1944.

Div. 4-232. 2-M20

17. Specification of Power Supply PS-1 and PS-2,

NBS, Ordnance Development Division, Dec. 9,
1944. Div. 4-232.2-21

NDRC ENGINEERING REPORTS

18. Status Report on Design of Generator-Powered
Radio Fuze, Chester H. Page and F. Stanley
Atchison, Service Projects OD-27 and SC-40,
Engineering Report 1-R, May 29, 1943.

Div. 4-211.2-M2

19. Status of Generator Development, George V.
Morris, Service Project OD-27, Engineering Re-
port 3-R, Zenith Radio Corporation, May 27, 1943.

Div. 4-232. 2-M6

20. Preliminary Discussion of Amplifier Simplification
For MC-382 Fuze, R. H. Pintell, Service Project
OD-27, Memorandum Report 35-R, Emerson Radio
and Phonograph Corporation, Apr. 8, 1943.

Div. 4-238. 222-M2

21. Status of Generator Development, R. N. Harmon,

Service Project OD-27, Memorandum Report 38-R,
Westinghouse Electric and Manufacturing Com-
pany, Apr. 8, 1943. Div. 4-232.2-M3

22. Generator Regulation, Chester H. Page, Service

Project OD-27, Memorandum Report 40-R, Apr.
26, 1943. Div. 4-232.2-M4

23. Amplifier Specifications for MC-382 Fuze, R. H.

Pintell, Service Project OD-27, Parts I and II,
Memorandum Report 48-R, Emerson Radio and
Phonograph Corporation, May 24 and July 24,
1943. Div. 4-238.222-M3

REPORT OF ORDNANCE DEVELOPMENT
DIVISION OF NATIONAL BUREAU
OF STANDARDS

24. Leakage Resistance of BS-U and BS-5 Detonators,
W. A. Yates, Report OD-1-75, NBS, Ordnance De-
velopment Division, Dec. 4, 1943.

Div. 4-238.523-MI

BIBLIOGRAPHY 443

25. BS-4 Detonators Fired through Sylvania SA-782-B
Thyratrons, Summary Report on Recent Tests, G.
Singer and T. N. White, Report OD-1-82, NBS,
Ordnance Development Division, Dec. 21, 1943.

Div. 4-238. 521-M6

26. Time Lags in BS-1* Detonators When Fired with-

out Firing Condensers, L. C. Miller, Report OD-1-
154, NBS, Ordnance Development Division, Feb.
15, 1944. Div. 4-238. 521-M8

27. BRLG Generator Speeds for Several Combinations
of Vehicle, Propeller Lead, and Manufacturer,
D. C. Friedman, Report OD-1-256 and 256A
(Supplement), NBS, Ordnance Development Di-
vision, May 22, 1944 and June 6, 1944.

Div. 4-232. 2-M14 (Supp.), Div. 4-232.2-M13

28. Field Test — 27 Philco T50E1 with Metal Pro-
pellers (PX-5), D. A. Worcester and D. C. Fried-
man, Report OD-1-405, NBS, Ordnance Develop-
ment Division, July 17, 1944. Div. 4-222. 111-MI

29. Field Test — 1*0 Bowen T50E10 Units Lot 11*1 (CB-

1*20), E. F. Horton and R. Vorkink, Report OD-1-
585, NBS, Ordnance Development Division, Dec.
14, 1944. Div. 4-222. 111-M2

30. Field Test — 21 Zenith T51 Units, Lot 53, (CB-

1*30), D. A. Worcester and G. Rabinow, Report
OD-1-626, NBS, Ordnance Development Division,
Jan. 19, 1945. Div. 4-222.112-MI

31. Field Test — Philco T50E1 Reporters with Dough-

nut Arming Ring, D. W. Scott, Report OD-1-660,
NBS, Ordnance Development Division, Feb. 22,
1945. Div. 4-222. 127-M2

32. Lot Quality Test of 12 Philco T-30 Units (TBG-

95), R. G. Tobey and G. Rabinow, Report OD-1-
664, NBS, Ordnance Development Division, Mar. 2,
1945. Div. 4-222. 124-M2

33. High Altitude Test, 21* Zenith T-51 Units (CB-

1*57), D. A. Worcester and G. Rabinow, Report

OD-1-684, NBS, Ordnance Development Division,
Mar. 24, 1945. Div. 4-222.112-M2

34. BS-5 Detonators Fired with 1.5 Microfarad Con-
denser, Charles C. Gordon, Report OD-1-699, NBS,
Ordnance Development Division, Apr. 2, 1945.

Div. 4-238.522-M3

35. High Altitude Test — 12 Zenith T-51 Units (CB-

1*61*), D. A. Worcester and G. Rabinow, Report
OD-1-701, NBS, Ordnance Development Division,
Apr. 12, 1945. Div. 4-222.112-M3

36. Arming Test — 18 Westinghouse T-82 Units (CB-

1*71*), D. A. Worcester and G. Rabinow, Report
OD-1-715, NBS, Ordnance Development Division,
Apr. 19, 1945. Div. 4-222.113-MI

37. Field Test — 20 Westinghouse T-82 Units (CB-

1*73), D. A. Worcester and R. Vorkink, Report
OD-1-733, NBS, Ordnance Development Division,
May 8, 1945. Div. 4-222.113-M3

38. Field Test — 20 Westinghouse T-82 Units (CB-

1*79), D. A. Worcester and R. Vorkink, Report
OD-1-736, NBS, Ordnance Development Division,
May 8, 1945. Div. 4-222.113-M4

39. Field Test — 21 Zenith T-51 Units (CB-1*81), R.
Vorkink, Report OD-1-749, NBS, Ordnance De-
velopment Division, May 18, 1945.

Div. 4-222. 112-M4

40. Field Test — 18 Emerson T-92 Units (CB-1*82),

D. A. Worcester and R. Vorkink, Report OD-1-755,
NBS, Ordnance Development Division, May 21,
1945. Div. 4-222. 114-MI

41. 73 — Globe Union T-132 (CHP-1*3) , R. G. Tobey
and D. C. Friedman, Report OD-1-763, NBS,
Ordnance Development Division, June 4, 1945.

Div. 4-222.131-M3

42. Reporter Test — 10 Westinghouse T-82E1 Units

(BX-12), D. A. Worcester and G. Rabinow, Report
OD-1-879, NBS, Ordnance Development Division,
Aug. 28, 1945. Div. 4-222.113-M5

43. A Two-Stage Feedback Amplifier, Ralph Stair,
Thomas M. Marion, and E. Eisner, Report OD-2-6,
NBS, Ordnance Development Division, Nov. 24,

1943. Div. 4-238. 227-MI

44. Regulation with Non-Linear Resistors in Series
with Load Current, J. G. Hoffman, Report OD-2-7,
NBS, Ordnance Development Division, Jan. 1,

1944. Div. 4-236-M2

45. Investigation of Design Features of Westinghouse
MK Generators, J. G. Reid, Jr., and Charles
Ravitsky, Report OD-2-20, NBS, Ordnance De-
velopment Division, Feb. 23, 1944.

Div. 4-232. 2-M12

46. Two-Tube Amplifier for BRTG-Pl*B Audio Am-

plifier, Ralph Stair and Thomas M. Marion, Report
OD-2-33, NBS, Ordnance Development Division,
May 13, 1944. Div. 4-238.226-MI

47. BRTG-P1*C Amplifier, Ralph Stair and Thomas M.
Marion, Report OD-2-38, NBS, Ordnance De-
velopment Division, June 7, 1944.

Div. 4-238.226-M2

48. Arming of VT Bomb Fuzes, A. L. Leiner, Report

OD-2-275, NBS, Ordnance Development Division,
Sept. 15, 1945. Div. 4-244.1-M3

49. Pentode Acceptance Amplifier, Robert D. Huntoon,

Report OD-BE-1, NBS, Ordnance Development
Division, June 19, 1944. Div. 4-238.227-M3

50. Comparison of Radiated Power of OD and RGD

Oscillators, R. F. Morrison, Jr., Report OD-BE-
7R, NBS, Ordnance Development Division, July
17, 1944. Div. 4-238.3-M2

51. Comparison of OD and RGD Circuits, R. B.
Schwartz, Report OD-BE-13R, NBS, Ordnance
Development Division, July 29, 1944.

Div. 4-238.3-M3

52. Voltage Relationships in the RGD Oscillator, R. F.
Morrison, Jr., Report OD-BE-30R, NBS, Ordnance
Development Division, Aug. 23, 1944.

Div. 4-238. 32-M9

53. Measurement of BRTG Sensitivity , R. F. Morrison,

Jr., Report OD-BE-39R, NBS, Ordnance Develop-
ment Division, Sept. 18, 1944. Div. 4-625-M2

54. ^ Delay of 10-E Amplifier, R. B. Schwartz, Report

444

BIBLIOGRAPHY

OD-BE-41R, NBS, Ordnance Development Di-
vision, Sept. 25, 1944. Div. 4-238.223-M3

55. A High Gain Amplifier Employing a Twin Triode

Tube, Thomas M. Marion, Report OD-BE-47R,
NBS, Ordnance Development Division, Oct. 18,
1944. Div. 4-231.3-M3

56. Probability Distribution of Arming Time Using

RC Arming, Charles Ravitsky, Report OD-BE-
49R, NBS, Ordnance Development Division, Oct.
23, 1944. Div. 4-238.514-M3

57. Analysis of the BS4 Detonator, Charles Ravitsky,

Report OD-BE-73R, NBS, Ordnance Development
Division, Mar. 7, 1945. Div. 4-238. 521-M9

58. The Detonator Circuit, Charles Ravitsky, Report

OD-BE-74R, NBS, Ordnance Development Di-
vision, Mar. 7, 1945. Div. 4-238. 523-M4

59. Measurement of Firing Voltage, Robert D.

Huntoon, NBS, Ordnance Development Division,
Aug. 20, 1943. Div. 4-621-M2

60. Gain Control for Amplifiers, Robert D. Huntoon

and F. Lamar Cooke, NBS, Ordnance Development
Division, Aug. 23, 1943. Div. 4-238.211-MI

61. Preliminary Information on Audio Amplifier for

BRLG-10, Robert D. Huntoon and F. Lamar
Cooke, NBS, Ordnance Development Division,
Sept. 18, 1943. Div. 4-238.222-M4

62. Status Report on Rectifier Sub Group, F. Stanley
Atchison, Report OD-3-I, NBS, Ordnance De-
velopment Division, Aug. 11, 1943.

Div. 4-235-MI

63. Tolerances on Complete BRLG-8, Robert D.
Huntoon, Report OD-3-6a, NBS, Ordnance De-
velopment Division, Oct. 22, 1943. Div. 4-211. 21-Ml

64. Performance of Westinghouse AQ Copper Oxide

Rectifying Cells, F. Stanley Atchison, Report OD-
3-VII, NBS, Ordnance Development Division, Aug.
24,1943. Div. 4-235-M2

65. Effect of Static Characteristics of Rectifier Cells

on A and B Voltages, F. Stanley Atchison, Report
OD-3-IX, NBS, Ordnance Development Division,
Sept. 15, 1943. Div. 4-235-M3

66. Critical Grid Voltage of Thyratron and Hum

Voltage Output of BRLG-11, F. Lamar Cooke, Re-
port OD-3-9, NBS, Ordnance Development Di-
vision, Oct. 27, 1943. Div. 4-231. 1-M7

67. Methods of Measuring the Critical Voltage of

Thyratrons, F. Lamar Cooke, Report OD-3-13,
NBS, Ordnance Development Division, Nov. 9,
1943. Div. 4-231. 1-M8

68. Generator Performance, William L. Kraushaar,

Report OD-3-17, NBS, Ordnance Development
Division, Nov. 1, 1943. Div. 4-232.2-M8

69. Performance of Power Supply at High and Low

Temperatures, F. Stanley Atchison, Report OD-
3-23, NBS, Ordnance Development Division, Nov.
6, 1943. Div. 4-232.2-M9

70. BRLG-11 A Amplifier for Zell Manufacture, Robert
D. Huntoon, Report OD-3-24, NBS, Ordnance De-
velopment Division, Nov. 8, 1943.

Div. 4-238.224'-Ml

71. Amplifier Performance of BRLG-8 Potted with
Glidden Compound, Albert Weiss, Report OD-3-26,
Nov. 8, 1943, and OD-3-26a, Nov. 18, 1943, NBS,
Ordnance Development Division, Nov. 8, 1943.

Div. 4-239.1-M4

72. Experiments with Standard MC-382 Fuzes Con-

verted to Reaction Type Fuzes with Grid Detection
(RGD Fuze), Philip Krupen and W. S. Hinman,
Jr., Report OD-3-27, NBS, Ordnance Development
Division, Nov. 15, 1943. Div. 4-238. 32-MI

73. Discussion of Proposed Rectifier Specifications,
F. Stanley Atchison, Report OD-3-28, NBS, Ord-
nance Development Division, Nov. 6, 1943.

Div. 4-235-M5

74. Preliminary Report on Tuning and Loading De-

vices for BRLG, Paul E. Landis, Report OD-3-37,
NBS, Ordnance Development Division, Nov. 29,
1943. Div. 4-233.1-MI

75. New Amplifier Design, Robert D. Huntoon, Re-

port OD-3-38, NBS, Ordnance Development Di-
vision, Nov. 29, 1943. Div. 4-238.213-MI

76. Component Specifications for BRLG-11 A, Robert
D. Huntoon, Report OD-3-39, NBS, Ordnance De-
velopment Division, Dec. 2, 1943.

Div. 4-238. 224-M2

77. Design and Tolerance Curves for BRLG-11 A, F.

Lamar Cooke and Robert D. Huntoon, Report
OD-3-40, NBS, Ordnance Development Division,
Dec. 3, 1943. Div. 4-238.224-M3

78. Effect of Component Tolerances on Performance

of BRLG-11 A, Robert D. Huntoon, Report OD-

3-46, NBS, Ordnance Development Division, Dec.
7, 1943. Div. 4-238. 224-M4

79. Report on Status of Work on RGD, Bertrand J.
Miller, Report OD-3-47, NBS, Ordnance Develop-
ment Division, Dec. 7, 1943. Div. 4-238.32-M2

80. Experiments with the RGD Circuit, Applied to

BRLG-8, William L. Kraushaar, Report OD-3-48,
NBS, Ordnance Development Division, Dec. 9,
1943. Div. 4-238.32-M3

81. Status of BRLG Production Designs, W. S. Hin-
man, Jr., Report OD-3-57, NBS, Ordnance De-
velopment Division, Dec. 16, 1943.

Div. 4-211. 21-M3

82. Effect of Tolerances in the Regulation Networ'k,
William L. Kraushaar, Report OD-3-60, NBS,
Ordnance Development Division, Dec. 17, 1943.

Div. 4-232. 2-M10

83. Performance of Zell 11 A Amplifiers on Standard

Test Voltages, Robert D. Huntoon, Report OD-3-63,
NBS, Ordnance Development Division, Dec. 23,
1943. Div. 4-238. 224-M5

84. Arming Considerations in T6, Bertrand J. Miller
and Philip R. Karr, Report OD-3-74, NBS, Ord-
nance Development Division, Jan. 22, 1944.

Div. 4-238.515-MI

85. An RGD Circuit for the MC-382, Philip Krupen,

Report OD-3-79, NBS, Ordnance Development Di-
vision, Jan. 15, 1944. Div. 4-238.32-M4

86. BRLG-10 A, F. Lamar Cooke, Report OD-3-94,

BIBLIOGRAPHY

445

NBS, Ordnance Development Division, Feb. 3,
1944. Div. 4-238.227-M2

87. Arming Resistor for T5, Robert D. Huntoon, Re-

port OD-3-101, NBS, Ordnance Development Di-
vision, Feb. 22, 1944. Div. 4-236-M3

88. RGD Circuit for BRLG Applications, Philip
Krupen, Report OD-3-102, NBS, Ordnance De-
velopment Division, Feb. 24, 1944.

Div. 4-238.32-M5

89. Amplifier Characteristics for T6 Application,

Charles J. Apolenis and Robert D. Huntoon, Re-
port OD-3-107, NBS, Ordnance Development Di-
vision, Mar. 7, 1944. Div. 4-238. 225-MI

90. Microphonic Stability of the Oscillator-Diode Type

of Fuze Circuit, Robert D. Huntoon, Report OD-
3-117, NBS, Ordnance Development Division, Mar.
22, 1944. Div. 4-238.31-MI

91. Dummy Antennas, Robert D. Huntoon, Report OD-

3-133, NBS, Ordnance Development Division, Apr.
20, 1944. Div. 4-233-M3

92. MRLG Apex Firing and Generator Regulation,
Chester H. Page, Report OD-3-142, NBS, Ord-
nance Development Division, May 9, 1944.

Div. 4-512-MI

93. Linearity of 11 A Amplifier, George Nordquist, Re-

port OD-3-148, NBS, Ordnance Development Di-
vision, May 13, 1944. Div. 4-238.224-M6

94. Uniformity of Raytheon Triodes in RGD, Chester
H. Page, Report OD-3-149, NBS, Ordnance De-
velopment Division, May 13, 1944. Div. 4-231. 3-MI

95. Triode Microphonics, Robert D. Huntoon, Bert-

rand J. Miller, and R. B. Schwartz, Report OD-3-
153, NBS, Ordnance Development Division, May
20, 1944. Div. 4-231.3-M2

96. Behavior of the 11 A Amplifier at 5,000 CPS, Philip

R. Karr and George Nordquist; Report OD-3-156,
NBS, Ordnance Development Division, May 25,
1944. Div. 4-238. 224-M7

97. Amplifier Hum Suppression, Robert D. Huntoon,
and Philip R. Karr, Report OD-3-158, NBS, Ord-
nance Development Division, June 9, 1944.

Div. 4-238.213-M2

98. Voltage Speed Regulation of Zenith Generators,

Morris Brenner and Ralph L. Ueberall, Report
OD-3-167, NBS, Ordnance Development Division,
July 1, 1944. Div. 4-232.2-M15

99. Effect of Low Temperature and High Voltage on
Performance of 11 A Amplifier, Philip R. Karr
and Milton Weiss, Report OD-3-169, NBS, Ord-
nance Development Division, July 19, 1944.

Div. 4-238.224-M8

100. 10E Amplifier, Philip R. Karr and Chester H.
Page, Report OD-3-170, NBS, Ordnance Develop-
ment Division, July 21, 1944. Div. 4-238.223-MI

101. Effect of Amplifier Shape on Function Height of
T50 E10, Philip R. Karr, Report OD-3-172, NBS,
Ordnance Development Division, Aug. 11, 1944.

Div. 4-238.223-M2

102. Effect of Potting Upon Amplifier Shaping, Philip
R. Karr and George Nordquist, Report OD-3-175,

NBS, Ordnance Development Division, Aug. 17,
1944. Div. 4-239.1-M5

103. Component Tolerance Study on BRTG-P5 Am-

plifier, Chris Gregory and Ralph Stair, Report
OD-3-180, NBS, Ordnance Development Division,
Sept. 22, 1944. Div. 4-238.226-M3

104. Effect of Low Plate Supply Voltage on RGD-PB

Units, Philip Krupen and Leonard C. Pochop, Re-
port OD-3-184, NBS, Ordnance Development Di-
vision, Oct. 6, 1944. Div. 4-238.32-M10

105. Universal High Gain Amplifier, George Nordquist,

Report OD-3-186, NBS, Ordnance Development

Division, Oct. 20, 1944. Div. 4-238.211-M3

106. The Performance of Zenith BRTG-Z Units as
Function of Supply Voltages, Lawrence J. Diou
and Ralph Stair, Report OD-3-189, NBS, Ord-
nance Development Division, Oct. 27, 1944.

Div. 4-621-M3

107. Tube and Component Study of 10E Amplifier,
Chris Gregory, Report OD-3-190, NBS, Ordnance
Development Division, Oct. 30, 1944.

Div. 4-238. 223-M4

108. Test of Four Types of Power Supplies and Gen-
erators (Quam Nichols, Utah, Knapp-Monarch
and Wurlitzer), Ralph L. Ueberall, Report OD-3-

193, NBS, Ordnance Development Division, Nov.

9, 1944. Div. 4-232.2-M18

109. Experimental Production of High Gain Modified
White Amplifiers, Philip R. Karr, Report OD-3-

194, NBS, Ordnance Development Division, Nov.

8, 1944. Div. 4-238.211-M4

110. Use of Off -Tolerance Condensers in the 10E Am-
plifier, George Nordquist, Report OD-3-195, NBS,
Ordnance Development Division, Nov. 11, 1944.

Div. 4-237-M5

111. Sensitivity of BRTG-POD, Glenn L. Scillian and
Chester H. Page, Report OD-3-199, NBS, Ord-
nance Development Division, Nov. 17, 1944.

Div. 4-238.31-M2

112. Alteration of Feedback Components in the Basic

10E Circuit, George Nordquist, Report OD-3-200,
NBS, Ordnance Development Division, Nov. 18,
1944. Div. 4-238.223-M5

113. Electrical Design Considerations for T-30, William

E. Kraushaar, R. B. Schwartz, and Bertrand J.
Miller, Report OD-3-203, NBS, Ordnance Develop-
ment Division, Dec. 5, 1944. Div. 4-241-M4

114. The BRTG-T1B Amplifier, Ralph Stair and Glenn
L. Scillian, Report OD-3-204, NBS, Ordnance De-
velopment Division, Dec. 7, 1944.

Div. 4-238.225-M2

115. Proposed Amplifier for T-30 (Air-to-Ground, Air-
to-Air), Philip R. Karr, Report OD-3-205, NBS,
Ordnance Development Division, Dec. 12, 1944.

Div. 4-238. 225-M3

116. The 11-N2 Medium Band Amplifier, George Nord-
quist, Report OD-3-208, NBS, Ordnance Develop-
ment Division, Jan. 8, 1945. Div. 4-238. 227-M4

117. Plate Voltage Fluctuations of Generator Power
Supplies, James H. Barnard, Leonard C. Pochop,

SECRET

446

BIBLIOGRAPHY

and Ralph Stair, Report OD-3-210, NBS, Ord-
nance Development Division, Jan. 22, 1945.

Div. 4-232. 2-M22

118. An RGD Oscillator for Working into High Radia-

tion Resistances , Richard F. Mills, Report OD-3-
212, NBS, Ordnance Development Division, Jan.
24, 1945. Div. 4-238.32-M11

119. Use of T-30 Filter with Wurlitzer Generator,
Chester H. Page, Report OD-3-214, NBS, Ord-
nance Development Division, Jan. 31, 1945.

Div. 4-237-M6

120. T-50 Function Height for Various Amplifiers

under Manifold Release Conditions, Mary L. Scott,
Report OD-3-215, NBS, Ordnance Development

Division, Feb. 2, 1945. Div. 4-241-M5

121. A Simplified RGD-PB Oscillator, Paul Miller, Re-

port OD-3-216, NBS, Ordnance Development Di-
vision, Feb. 7, 1945. Div. 4-238.32-M12

122. Revised Circuit for BRTG-TlB Amplifier, Dorothy

R. Adams and George Nordquist, Report OD-3-
219, NBS, Ordnance Development Division, Mar.
2, 1945. Div. 4-238.225-M5

123. The T-132 (mortar fuze) Apex Performance Prob-

lem, William L. Kraushaar, Report OD-3-220,
NBS, Ordnance Development Division, Mar. 3,
1945. Div. 4-222.131-M2

124. Selection of Screen Grid Voltage Divider for

MRLG-1, T-132 , in Connection with the Apex Fir-
ing Problem, George Nordquist, Report OD-3-221,
NBS, Ordnance Development Division, Mar. 10,
1945. Div. 4-512-M2

125. Temperature Coefficient of Allen-Bradley , Erie,

Continental Carbon % Watt Resistors, and IRC
XA Watt Resistors, F. W. Jirauch, Report OD-3-
222, NBS, Ordnance Development Division, Mar.
2, 1945. Div. 4-236-M11

126. Frequency Modulation in Generators, Ralph Stair
and Glenn L. Scillian, Report OD-3-223P, NBS,
Ordnance Development Division, Mar. 12, 1945.

Div. 4-232. 2-M23

127. Suggested Amplifier for T-132L Having Low Gain

at Low Frequency, George Nordquist, Report OD-
3-225, NBS, Ordnance Development Division, Mar.
12, 1945. Div. 4-238.225-M6

127a. The RGD Oscillator, Philip Krupen, Report

OD-3-227, NBS, Ordnance Development Di-
vision, Mar. 14, 1945. Div. 4-238.32-M13

128. The Effect of Tube Parameters on the Available
Gain of Amplifiers, Chris Gregory, Report OD-3-

231, NBS, Ordnance Development Division, Mar.

17, 1945. Div. 4-238.211-M5

129. A Quasi-Hartley Plate-Loaded RGD Oscillator,

Paul Miller and Richard F. Mills, Report OD-3-

232, NBS, Ordnance Development Division, Mar.

20, 1945. Div. 4-238.32-M14

130. Preliminary Report on Heights of Function with

Proposed Universal Amplifier for Mortar Applica-
tion, Philip R. Karr, Mary L. Scott, and George

Nordquist, Report OD-3-235P, NBS, Ordnance De-
velopment Division, Apr. 4, 1945. Div. 4-241-M6

131. Addendum to OD-3-235P, George Nordquist, April

16, 1945. Div. 4-241-M6

132. Comparison of Wire-Wound and Ceramic Gain

Controls for Use in the 10E Amplifier, F. W.
Jirauch and Donald G. Green, Preliminary Re-
port OD-3-236P, NBS, Ordnance Development Di-
vision, Apr. 7, 1945. Div. 4-238.211-M6

133. Apex Performance of the T-171 Mortar Fuze with

RC Arming Delay, Philip Krupen, Report OD-3-
242, NBS, Ordnance Development Division, May
5, 1945. Div. 4-238.514-M4

134. Thyratron Normal Critical Voltages for Various

Amplifiers, George Nordquist, Report OD-3-249,
NBS, Ordnance Development Division, May 24,
1945. Div. 4-231. 1-M11

135. Temperature Effect on G-U T-132 Amplifiers and
Amplifier Components, F. W. Jirauch and Donald
G. Green, Preliminary Report OD-3-252P, NBS,
Ordnance Development Division, May 28, 1945.

Div. 4-238.225-M8

136. Response of Shaped Amplifiers to Step Pulses and
Sharp Pulses, Philip R. Karr, R. B. Schwartz, and
Mary L. Scott, Report OD-3-253, NBS, Ordnance
Development Division, May 31, 1945.

Div. 4-238.212-M4

137. Pentode Input Impedance, George Nordquist, Re-

port OD-3-254, NBS, Ordnance Development Di-
vision, May 31, 1945. Div. 4-231.4-M12

138. Temperature Characteristics of the Ceramic Con-
densers in the Globe-Union T-132 Amplifier, F. W.
Jirauch, Report OD-3-255P, NBS, Ordnance De-
velopment Division, June 4, 1945. Div. 4-237-M9

139. Grid Bias Circuit for T-171 Mortar Fuze to Re-
duce Apex Malfunction, George Nordquist and
Dorothy R. Adams, Report OD-3-256, NBS, Ord-
nance Development Division, June 5, 1945.

Div. 4-238.514-M5

140. Transparent Charts for Prediction of Function

Height, Philip R. Karr, Chris Gregory, R. B.
Schwartz, and Mary L. Scott, Report OD-3-257,
NBS, Ordnance Development Division, June 6,
1945. Div. 4-241-M7

141. Revised T-2005 Amplifier, Dorothy R. Adams, Re-

port OD-3-264, NBS, Ordnance Development Di-
vision, July 30, 1945. Div. 4-238.225-M9

142. A Study of Some Amplifier Curves for Use with
the MU3C Mortar, Mary L. Scott and George Nord-
quist, Report OD-3-267P, NBS, Ordnance De-
velopment Division, July 4, 1945.

Div. 4-238. 213-M3

143. Correlation of Rotor Magnetic Characteristics
with Generator Output, Glenn L. Scillian and
Ralph L. Ueberall, Report OD-3-269, NBS, Ord-
nance Development Division, July 5, 1945.

Div. 4-232. 22-M9

144. Effect of Key Components on Amplifier Response
Characteristics, George Nordquist, Report OD-3-

BIBLIOGRAPHY

447

275, NBS, Ordnance Development Division, July
16, 1945. Div. 4-238.213-M4

145. Air Speed-Generator Output Regulation for Mor-
tar Shell Fuzes , Glenn L. Scillian and L. M. An-
drews, Report OD-3-278, NBS, Ordnance Develop-
ment Division, July 20, 1945. Div. 4-232. 2-M25

146. Arming Pulse Protection Circuit. Philip R. Karr,

William L. Kraushaar, and Chester H. Page, Re-
port OD-3-284, NBS, Ordnance Development Divi-
sion, Sept. 14, 1945. Div. 4-238.515-M6

147. Effect of Different Regulation Networks on T-132

Generator Speeds, Glenn L. Scillian, Report OD-
3-286, NBS, Ordnance Development Division, Sept.
14, 1945. Div. 4-232. 2-M26

148. Generator Voltage Measurements, F. Manov and

Jacob Rabinow, Report OD-4-1, NBS, Ordnance
Development Division, Aug. 13, 1943.

Div. 4-621-MI

149. Absolute Frictional Torque of Generator Bearings,
A. Chartock and L. B. Heilprin, Report OD-4-7,
NBS, Ordnance Development Division, Nov. 29,

1943. Div. 4-232.23-MI

150. Speed Regulating Propellers, Jacob Rabinow, Re-

port OD-4-11, NBS, Ordnance Development Divi-
sion, Dec. 4, 1943. Div. 4-232.21-MI

151. Specification of Maximum Starting Torque of

Complete BRLG Unit, A. Chartock and L. B. Heil-
prin, Report OD-4-13, NBS, Ordnance Develop-
ment Division, Dec. 6, 1943. Div. 4-211.21-M2

152. Propeller Torque at Low Velocity, L. M. Andrews,

Report OD-4-19, NBS, Ordnance Development Di-
vision, Dec. 21, 1945. Div. 4-232.21-M3

153. Bursting Speed of Generator Rotors, Samuel
Kolodny, Report OD-4-6, NBS, Ordnance Develop-
ment Division, Feb. 23, 1944. Div. 4-232.22-M4

154. Measurement of Dynamic Propeller Unbalance,
E. U. Rotor and L. G. Koontz, Report OD-4-43,
NBS, Ordnance Development Division, Mar. 23,

1944. Div. 4-232. 21-M5

155. Comparative Speeds of Brass and Bakelite Pro-

pellers (Supp. 1), Louis Schuman, Report OD-4-45,
NBS, Ordnance Development Division, Apr. 6,
1944. Div. 4-232. 21-M6

156. Propeller Unbalance Tester (3 supplements),
Jacob Rabinow, Report OD-4-48, NBS, Ordnance
Development Division, Apr. 20, 1944.

Div. 4-616-MI

157. Torque Developed by 2"xl2" RRLG Propellers,

Samuel Kolodny, Report OD-4-51, NBS, Ordnance
Development Division, Apr. 19, 1944.

Div. 4-232. 21-M7

158. Speed Tests of Stamped Brass and Duralumin

Propellers (3 supplements), Louis Schuman, Re-
port OD-4-60, NBS, Ordnance Development Divi-
sion, May 22, 1944. Div. 4-232.21-M8

159. Life Test on Oilite Bearings of MRLG Units,
A. Chartock, Report OD-4-74, NBS, Ordnance De-
velopment Division, June 16, 1944.

Div. 4-232. 23-M3

160. Determination of Static Thrust Load Limit for the
V2 Inch New Departure R-3 Ball Bearing, A. Char-
tock, Report OD-4-83, NBS, Ordnance Develop-
ment Division, Sept. 19, 1944. Div. 4-232. 23-M4

161. The Use of Precision Bearings in BRLG and T-50
Noses, Jacob Rabinow, Report OD-4-88, NBS, Ord-
nance Development Division, Dec. 14, 1944.

Div. 4-232.23-M6

162. Fifty Indiana Steel and Arnold Engineering
Rotors Submitted for Test by Bowen and Com-
pany, Samuel Kolodny, Report OD-4-97, NBS,
Ordnance Development Division, Mar. 2, 1945.

Div. 4-232.22-M6

163. Effect of Condenser Leakage on RC Arming, Cledo
Brunetti, Report OD-5-126, NBS, Ordnance De-
velopment Division, Sept. 23, 1943.

Div. 4-238.514-MI

164. Noise Performance of Raytheon Diodes, M. Schul-

kin, Report OD-5-224, NBS, Ordnance Develop-
ment Division, Dec. 7, 1943. Div. 4-231. 2-MI

165. Surge Current Performance and Requireme?its of

BRLG Filter Condensers, Willis E. Armstrong,
Report OD-5-594, NBS, Ordnance Development
Division, Sept. 13, 1944. Div. 4-237-M3

366. Minimum Capacity Requirements for the BRLG
Filter Condensers, Willis E. Armstrong, Report
OD-5-655, NBS, Ordnance Development Division,
Oct. 10, 1944. Div. 4-237-M4

167. Revised Amplifier for T-91, Paul E. Landis and
George Nordquist, Report OD-5-765, NBS, Ord-
nance Development Division, Mar. 29, 1945.

Div. 4-238. 225-M7

168. Change in T-91 Amplifier to Obtain a Longer
Trimmer Condenser, Cledo Brunetti and George
Nordquist, Report OD-5-769, NBS, Ordnance De-
velopment Division, Apr. 2, 1945. Div. 4-237-M8

169. RF Sensitivity for the Zenith T-172 Unit and

Variations Thereof, Otto E. Spokas, Report OD-
7-214R, NBS, Ordnance Development Division,
Aug. 13, 1945. Div. 4-233-M6

MEMORANDA OF ORDNANCE DEVELOPMENT
DIVISION OF NATIONAL BUREAU
OF STANDARDS

170. Tests on Reliability of Firing, Minimum Reliable
Firing Voltage and Time Lags for BS-5 Squibs,
W. A. Yates, Memorandum PG-319, NBS, Ord-
nance Development Division, Sept. 30, 1942.

Div. 4-238.522-MI

171. Specification Tests on BS-A Squibs, W. A. Yates,
Memorandum PG-380, NBS, Ordnance Develop-
ment Division, Nov. 6, 1942. Div. 4-238.521-Ml

172. Reliability of Firing BS-4 Squibs and Time Tests
with Radio Frequency Choke and with Resistance
in Series with Squib, Allen V. Astin and W. A.
Yates, Memorandum PG-383, NBS, Ordnance De-
velopment Division, Nov. 9, 1942.

Div. 4-238.521-M2

448

BIBLIOGRAPHY

173. Test on Minimum Firing Current of BS-U Squib,
W. A. Yates, Memorandum PG-395, NBS, Ord-
nance Development Division, Nov. 18, 1942.

Div. 4-238. 521-M3

174. Reliability of Firing and Time Test on SA782B,

Sylvania Thyratrons , W. A. Yates, Memorandum
PG-412, NBS, Ordnance Development Division,
Dec. 2, 1942. Div. 4-231.11-MI

175. Time Lag Specification for BS-4 Squibs, Allen V.
Astin and W. A. Yates, Service Project OD-27,
Memorandum Report 29-T, NBS, Ordnance De-
velopment Division, Jan. 29, 1943.

Div. 4-238.521-M4

176. Minimum Voltage to Fire BS-4 Detonators
through Thyratrons in Complete MC-382 Heads,
Memorandum from Theodore B. Godfrey to Harry
M. Diamond, by L. C. Miller, NBS, Ordnance De-
velopment Division, Dec. 30, 1943.

Div. 4-238. 521-M7

177. Firing Circuit Curves, Memorandum from Theo-

dore B. Godfrey to Messrs. Diamond, Astin, et al.,
NBS, Ordnance Development Division, July 26,
1944. Div. 4-238. 523-M2

178. Detonator Firing Test, Memorandum from W. A.

Yates to Cledo Brunetti, NBS, Ordnance Develop-
ment Division, Nov. 3, 1944. Div. 4-238.522-M2

179. Arming Considerations for HVAR, Bertrand J.
Miller, Memorandum OD-BE-17M, NBS, Ordnance
Development Division, July 12, 1944.

Div. 4-244.2-MI

180. Amplifier with Hum-Bucking for White RGD,
Philip R. Karr, Memorandum OD-3-28M, NBS,
Ordnance Development Division, Aug. 8, 1944.

Div. 4-238.32-M8

181. Incorporation of RC Arming for T-30, William L.
Kraushaar, Memorandum OD-3-48M, NBS, Ord-
nance Development Division, Oct. 20, 1944.

Div. 4-238. 514-M2

182. Temperature Coefficient of Condensers Used in the

10-E Amplifier, F. Jirauch, Memorandum OD-3-
86M, NBS, Ordnance Development Division, Feb.
24, 1945. Div. 4-237-M7

183. Modification of T-30 Amplifier, George Nordquist,
Memorandum OD-3-88M, NBS, Ordnance Develop-
ment Division, Mar. 2, 1945. Div. 4-238.225-M4

184. Possible Uses of Non-Linear Resistors, Philip
Krupen, Memorandum OD-3-99M, NBS, Ordnance
Development Division, May 15, 1945.

Div. 4-236-M12

185. Revision of Westinghouse T-82, Dorothy R.

Adams, George Nordquist, and Ralph Stair, Memo-
randum OD-3-122M, NBS, Ordnance Development
Division, Aug. 3, 1945. Div. 4-238. 225-M10

186. Report on Visit to Emerson Company, Memoran-

dum to W. S. Hinman, Jr., from Philip Krupen,
NBS, Ordnance Development Division, Apr. 22,
1943. Div. 4-238.7-MI

187. Study of Pintell Circuit, Memorandum to W. S.
Hinman, Jr., from Philip Krupen and F. L. Cooke,

NBS, Ordnance Development Division, Apr. 30,
1943. Div. 4-238.7-M2

188. Request for Laboratory Tests on Performance of

BS-A Squibs, Memorandum from Cledo Brunetti
to Theodore B. Godfrey, NBS, Ordnance Develop-
ment Division, Oct. 18, 1943. Div. 4-238. 521-M5

189. Relation of Thyratron Repeated Surge Perform-

ance to Time Delay, Abraham Silverstein, Report
OD-CT-M8, NBS, Ordnance Development Division,
Oct. 4, 1944. Div. 4-231.1-M9

REPORTS OF CONTRACTORS OF DIVISION U
OF NDRC

190. Summay'y of Activities in Development and Pilot

Manufacturing Run of Radio Fuses and Acces-
sories, Vernon D. Hauck, OEMsr-258, Friez In-
strument Division, Bendix Aviation Corporation,
Sept. 27, 1944. Div. 4-100-M4

191. Preliminary Draft of Final Technical Report

under Contracts OEMsr-885 and OEMsr-1113,
Emerson Radio and Phonograph Corporation, May
14, 1945. Div. 4-100-M6

192. Final Report of the Federal Telephone and Radio

Corporation, T. Smith Taylor, OEMsr-941, Fed-
eral Telephone and Radio Corporation, Oct. 5,
1943. Div. 4-235-M4

193. A Radio Proximity Fuze, Type BRTD (Part I),
OEMsr-949, University of Florida, Sept. 26, 1945.

Div. 4-211. 21-M13

194. Generator Powered Radio Proximity Fuze, Type
T-2005, Muriel E. Pottasch, OEMsr-1437, General
Instrument Corporation, Aug. 1, 1945.

Div. 4-222.125-M2

195. Generator Powered Radio Proximity Fuze for

Mortars, Longitudinal Excitation Type, Alfred S.
Khouri, OEMsr-1117, Globe-Union, Inc., Sept. 30,
1945. ' Div. 4-222. 131-M6

196. Summay'y Technical Report for Contract OEMsr-

769, OEMsr-769, State University of Iowa, Sept.
29, 1945. Div. 4-100-M7

197. [Alnico Rotor Generators], Final Report oyi Con-
tract OEMsr-113U, C. W. Clemons, OEMsr-1134,
Knapp-Monarch Company, Nov. 20, 1944.

Div. 4-232. 2-M19

198. [Generators], Final Report on NDRC Conti'act
OEMsr-981, C. W. Clemons, OEMsr-981, Knapp-
Monarch Company, Feb. 17, 1944.

Div. 4-232.2-M11

199. Final Report (Reserve Batteymy and Low Tempera-

ture Dry Cells), F. T. Bowditch and A. K. Hunt-
ley, OEMsr-528, National Carbon Company, Oct.
26, 1944. Div. 4-232.1-M9

199a. Final Report of Research and Development

Conducted by Philco Corporation on P4-772
Radio Proximity Fuze for Large Bombs,
R. A. Bell, OEMsr-866, Philco Radio and
Television Corporation, June 15, 1943.

Div. 4-211.1-M4

BIBLIOGRAPHY

449

200. Final Progress Report Contract OEMsr-1196,
Maurice E. Swift, OEMsr-1196, Philco Radio and
Television Corporation, May 31, 1945.

Div. 4-211.21-M12

201. [Vacuum Tubes, Types NR-2 (2B-24), NR-3

(2C-27) and NR-5 (2E-27)], Final Summary Re-
port [on] Contract OEMsr-566, A. Abate, OEMsr-
566, Raytheon Manufacturing Company, Oct. 1,
1945. Div. 4-231-M5

202. [Electronic Tubes], Final Report [on] Contract

OEMsr-630, OEMsr-630, Sylvania Electric Prod-
ucts, Inc., 1945. Div. 4-231-M4

203. [Battery Requirements for Project K-4], Final
Technical Report of Work Performed under
OEMsr-887, C. B. Pear, Jr., Washington Institute
of Technology, Feb. 17, 1944. Div. 4-232. 1-M8

204. Final Technical Report on Generator Powered

Proximity Fuzes for Bombs, Contract II, K. D.
Smith and A. L. Stillwell, OEMsr-905, Western
Electric Company and Bell Telephone Labora-
tories, May 30, 1944. Div. 4-211. 21-M5

205. Development of a Ground Approach Proximity
Fuze for Bomb Nose, BRTG, T. M. Bloomer,
OEMsr-343 and OEMsr-1106, Termination Report
CFE-760, Westinghouse Electric and Manufac-
turing Company, Apr. 28, 1945. Div. 4-211.21-Mll

206. Proximity Fuze, Bomb, Nose, Ground Approach;

Type VT, T-32, T. M. Bloomer, OEMsr-343 and
OEMsr-1106, Termination Report CFE-759, West-
inghouse Electric and Manufacturing Company,
Apr. 28, 1945. Div. 4-222.113-M2

207. RRLG Proximity Fuzes (Final Report), F. H.
Osborne, OEMsr-1161 and OEMsr-1163, Rudolph
Wurlitzer Company, Mar. 15, 1945.

Div. 4-211. 22-MI

208. Final Technical Report of Work Performed under

Contract OEMsr-95U, OEMsr-954, Zell Corpora-
tion, Jan. 12, 1945. Div. 4-100-M5

209. Mass Production of T-51 Fuzes by the Zenith

Radio Corporation, Earl J. Diehl, OEMsr-1477,
Service Project OD-27, Zenith Radio Corporation,
Oct. 30, 1945. Div. 4-222.112-M5

210. Generator-Powered Radio Proximity Fuze for
Mortars: Loop Transverse- Antenna Type, Earl
J. Diehl, OEMsr-1477, Service Project OD-27,
Zenith Radio Corporation, Oct. 30, 1945.

Div. 4-222. 132-MI

211. Development Report — 1 tV' Diameter Generator
for Fuze Well, George V. Morris, OEMsr-980,
Zenith Radio Corporation, Oct. 8, 1943.

Div. 4-232.2-M7

UNITED STATES MILITARY PUBLICATIONS

212. U. S. Army Ordnance Department Tentative Spec-
ification AXS-1199, February 10, 1944, Detonator,
Electric, T3.

213. U. S. Army Signal Corps Specification 371-2088,
Nov. 19, 1943, Electron Tube 2D29 (Thyratron).

UNCLASSIFIED TECHNICAL PUBLICATIONS

214. “An Alternating Current Dynamo with a Flat
Characteristic for Bicycle Illumination,” H. A. G.
Hazeu and M. Kiek, Phillips Technical Review,
Vol. Ill, No. 3, p. 87, March 1938.

215. Circular C448 of the National Bureau of Stand-
ards, “Permanent Magnets,” Raymond L. Sanford,
Aug. 10, 1944.

216. “Selenium Rectifier Characteristics, Application
and Design Factors,” C. A. Clarke, Electrical Com-
munication, Vol. 20, No. 1, 1941.

217. Radio Engineers Handbook , Frederick E. Ter-
man, McGraw-Hill Book Co., New York.

Chapter 4

REPORTS OF ORDNANCE DEVELOPMENT DIVI-
SION OF NATIONAL BUREAU OF STANDARDS

1. Six Speed Regulating Propellers on BRLG Self-

Reporters (Test Request WBM-9), Aberdeen, De-
cember 1, 19U3, D. C. Friedman, Report OD-1-76,
NBS, Ordnance Development Division, Dec. 11,
1943. Div. 4-232.21-M2

2. Six Speed Regulating Propellers on Self -Reporters
(WBM-10) , Aberdeen, January 23, 19 UU, D. C.
Friedman, Report OD-1-126, NBS, Ordnance De-
velopment Division, Jan. 31, 1944.

Div. 4-232.21-M4

3. Noise Produced by Qear Trains Using Various

Types of Planetary Gears, P. S. Manov, Report
OD-4-3, NBS, Ordnance Development Division,
Oct. 2, 1943. Div. 4-238.512-MI

4. Speed Regulating Propellers, Jacob Rabinow, Re-

port OD-4-11, NBS, Ordnance Development Divi-
sion, Dec. 4, 1943. Div. 4-232.21-MI

5. Comparison of Generator Rotor Unbalance and

the Measured Eccentricity, A. Donald Arsem, Re-
port OD-4-70, NBS, Ordnance Development Divi-
sion, Dec. 27, 1943. Div. 4-232.22-M3

6. Report on Mechanical Vibration of the BRLG

Units Mounted on M-6U Bomb, Jacob Rabinow,
Report OD-4-32, NBS, Ordnance Development
Division, Feb. 12, 1944. Div. 4-622-MI

7. SW-200 Switch Modified to Fire on Contact, Jacob
Rabinow, Report OD-4-44, NBS, Ordnance De-
velopment Division, Mar. 31, 1944.

Div. 4-238.511-M4

8. Propeller Unbalance Tester, Jacob Rabinow, Re-

port OD-4-48, NBS, Ordnance Development Divi-
sion, Apr. 20, 1944. Div. 4-616-MI

9. Equipment for Balancing Propellers, Jacob Rabi-

now and A. Donald Arsem, Report OD-4-48 Sup-
plement, NBS, Ordnance Development Division,

May 19, 1944. Div. 4-616-M2

10. Air Travel Required for Release of Arming Cover,
E. U. Rotor, Report OD-4-54, NBS, Ordnance
Development Division, Apr. 29, 1944.

Div. 4-244.1-MI

Turbine Dynamometer for Determining Input

450

BIBLIOGRAPHY

Torque of Gear Trains, Jacob Rabinow and Louis
Schuman, Report OD-4-63, NBS, Ordnance De-
velopment Division, May 10, 1944. Div. 4-612-Ml

12. Force Required to Pull Out Arming Wire on

BRLG Unit, Samuel Kolodny, Report OD-4-72,
NBS, Ordnance Development Division, June 13,
1944. Div. 4-238.513-Ml

13. Measurement of Vibration Amplitude of MRLG
Units, A. Chartock, Report OD-4-73, NBS, Ord-
nance Development Division, June 14, 1944.

Div. 4-232. 22-M5

14. Life Test on Oilite Bearings of MRLG Units,
A. Chartock, Report OD-4-74, NBS, Ordnance
Development Division, June 16, 1944.

Div. 4-232.23-M3

15. Improvements in the Arming System for the T-50
Fuze, Jacob Rabinow, Report OD-4-79, NBS, Ord-
nance Development Division, Aug. 23, 1944.

Div. 4-238. 513-M2

16. Effect of Generator End Play on Electrical Noise

Output, Louis Schuman and A. Donald Arsem,
Report OD-4-81, NBS, Ordnance Development Di-
vision, Sept. 7, 1944. Div. 4-232.2-M16

17. The Use of Precision Bearings in BRLG and T-50
Noses, Jacob Rabinow, Report OD-4-88, NBS, Ord-
nance Development Division, Dec. 14, 1944.

Div. 4-232. 23-M6

18. Effect of Varying Blade Length and Cover Open-

ings on Speed Characteristics and Air Thrust on
Turbine Wheel TFA6070, Louis Schuman, Report
OD-4-91, NBS, Ordnance Development Division,
Dec. 29, 1944. Div. 4-232.21-M14

19. Design of Impact Detonating Element for T-32
Fuze, Louis Schuman, Report OD-4-96, NBS, Ord-
nance Development Division, Feb. 17, 1945.

Div. 4-238. 523-M3

20. Supporting the T-132 and T-32 Generator to Take
Setback, Louis Schuman, Report OD-4-101, NBS,
Ordnance Development Division, Mar. 15, 1945.

Div. 4-232. 2-M24

21. Torsion Wire Dynamometer, Louis Schuman, Re-

port OD-4-105, NBS, Ordnance Development Divi-
sion, May 26, 1945. Div. 4-612-M2

22. Proposed Design for Dynamic Balancing Machine,
Jacob Rabinow, Report OD-4-108, NBS, Ordnance
Development Division, June 6, 1945.

Div. 4-616-M8

23. Method of Assembling Detonators to the T-132 /

T-171 Interrupter Rotors, Jacob Rabinow, Report
OD-4-124, NBS, Ordnance Development Division,
Aug. 3, 1945. Div. 4-238.523-M5

24. Second Test of Double -Element Setback Pins,
George T. Parish, Report OD-4-128, NBS, Ord-
nance Development Division, Sept. 5, 1945.

Div. 4-238. 513-M4

25. Nitrided Bearings, Ermo Furlani and Jacob
Rabinow, Report OD-4-132, NBS, Ordnance De-
velopment Division, Nov. 6, 1945.

Div. 4-232. 23-M9

MEMORANDA OF ORDNANCE DEVELOPMENT
DIVISION OF BUREAU OF STANDARDS

26. Setback Switches, Memorandum to Alexander

Ellett from William B. McLean, Jacob Rabinow,
and L. M. Andrews, NBS, Ordnance Development
Division, Mar. 9, 1942. Div. 4-238.511-Ml

27. Direction of Rotation of Escapement Wheel in
Setback Arming Devices, Memorandum to Alex-
ander Ellett from William B. McLean, NBS, Ord-
nance Development Division, Oct. 10, 1942.

Div. 4-238.511-M3

28. Rubber Mounted Generator Rotors, Memorandum

to Harry M. Diamond from William B. McLean,
NBS, Ordnance Development Division, Sept. 1,
1943. Div. 1-232.22-MI

29. Contact Springs in the BRLG Rotor Housing,

Memorandum to Harry M. Diamond from William
B. McLean, NBS, Ordnance Development Division,
Oct. 8, 1943. Div. 4-232.22-M2

30. Installation of Oilite Bearings in BRLG Genera-

tors, Memorandum to Harry M. Diamond from
William B. McLean, NBS, Ordnance Development
Division, May 4, 1944. Div. 4-232.23-M2

31. MRLG Gear Design, Memorandum to William B.
McLean from Jacob Rabinow, NBS, Ordnance De-
velopment Division, May 22, 1944.

Div. 4-238. 515-M3

32. Calibration of Propeller Unbalance Tester, Memo-

randum to William B. McLean from A. Donald
Arsem, NBS, Ordnance Development Division,
June 30, 1944. Div. 4-616-M3

33. Measurement of Gam of Balancing Equipment

(Propeller) , Memorandum to Jacob Rabinow from
A. Donald Arsem, NBS, Ordnance Development
Division, July 10, 1944. Div. 4-238.211-M2

34. Propeller Unbalance Specifications , Memorandum
to Harry M. Diamond from Jacob Rabinow, NBS,
Ordnance Development Division, Oct. 16, 1944.

Div. 4-616-M4

35. Metal Propeller with Fluted Blades, Memorandum
to Harry M. Diamond from Jacob Rabinow, NBS,
Ordnance Development Division, Nov. 1, 1944.

Div. 4-232. 21-M13

36. Coupling Shaft in Front Bearing Assemblies,

Memorandum to Harry M. Diamond from Jacob
Rabinow, NBS, Ordnance Development Division,
Nov. 13, 1944. Div. 4-232.23-M5

37. Visit to New Departure, January 5, 19 U5, Jacob
Rabinow, Memorandum OD-4-11M, NBS, Ord-
nance Development Division, Jan. 11, 1945.

Div. 4-232. 23-M7

38. Lock Washers, Jacob Rabinow, Memorandum OD-

4-12M, NBS, Ordnance Development Division, Jan.
12, 1945. Div. 4-239.2-MI

39. Some Comments of Field Personnel on Experience

with Bombs and Fuzes, Jacob Rabinow, Memo-
randum OD-4-19M, NBS, Ordnance Development
Division, Jan. 24, 1945. Div. 4-238.515-M4

40. Eliminating Noise Due to T-50 Gear Trains, Jacob

BIBLIOGRAPHY

451

Rabinow, Memorandum OD-4-21M, NBS, Ord-
nance Development Division, Feb. 7, 1945.

Div. 4-238.512-M2

41. Requirements for Doughnut Mechanism, Jacob

Rabinow and J. A. Senn, Memorandum OD-4-39M,
NBS, Ordnance Development Division, Mar. 17,
1945. Div. 4-238.515-M5

42. Arming Pin Considerations for the T-132, Jacob
Rabinow, Memorandum OD-4-44M, NBS, Ord-
nance Development Division, Apr. 7, 1944.

Div. 4-238.513-M3

43. Compilation of Performance of Various Rotors
Tested for Bursting Speed, Samuel Kolodny,
Memorandum OD-4-50M, NBS, Ordnance Develop-
ment Division, Apr. 28, 1945. Div. 4-232. 22-M7

44. Jolt Test of T-171 Bases, Louis Schuman, Memo-

randum OD-4-52M, NBS, Ordnance Development
Division, May 7, 1945. Div. 4-238.3-M4

45. Clock Rotor for the T-132 1 T-171, Jacob Rabinow,
Memorandum OD-4-67M, NBS, Ordnance Develop-
ment Division, June 21, 1945. Div. 4-232. 22-M8

REPORTS OF CONTRACTORS OF DIVISION 4
OF NDRC

46. Development of Balancing Equipment for T-171

Turbine Assembly, M. S. Redden and Allen S.
Clarke, OEMsr-1227, Bowen and Company, Elec-
tronics Division, May 1945. Div. 4-616-M7

47. Final Technical Report under Contracts OEMsr-
885 and OEMsr-1113, (Preliminary Draft),
OEMsr-885 and OEMsr-1113, Emerson Radio and
Phonograph Corporation, May 14, 1945.

Div. 4-100-M6

First Part of Final Report: Interim Reports 31
through 71, OEMsr-2163, Service Project P4-771R,
Emerson Radio and Phonograph Corporation.

48. Summary of Activities in Development and Pilot
Manufacturing Run of Radio Fuzes and Acces-
sories and Supplementary Report Covering De-
velopment of BRLG (Air Driven Alternator Prox-
imity Fuze), Final Report, Vernon D. Hauck,
OEMsr-258, Friez Instrument Division, Bendix
Aviation Corporation, Sept. 27, 1944.

Div. 4-100-M4

49. Generator Powered Proximity Fuze, Type T-2005,
Muriel E. Pottasch, OEMsr-1437, General Instru-
ment Corporation, Aug. 1, 1945.

Div. 4-222. 125-M2

50. Generator Powered Radio Proximity Fuze for

Mortars, Longitudinal Excitation Type T-132,
Alfred S. Khouri, OEMsr-1117, Globe-Union, Inc.,
Sept. 30, 1945. Div. 4-222.131-M6

51. Development and Manufacturing Report on NDRC
Gear Reduction Unit for VT Bomb Fuze, OEMsr-
1117, Globe-Union, Inc., Aug. 31, 1945.

Div. 4-238. 512-M4

52. [Alnico Rotor Generators], Final Report — Con-
tract OEMsr-1134, C. W. Clemons, OEMsr-1134,
Knapp-Monarch Company, Nov. 20, 1944.

Div. 4-232. 2-M19

53. Pilot Line Production of BRLG Equipment (Final

Progress Report), Maurice E. Swift, OEMsr-1196,
Philco Radio and Television Corporation, May 31,
1945. Div. 4-211. 21-M12

54. [The BRLG Unit], Final Report of the OSRD
Project, Olga E. Yeaton, OEMsr-866, Philco Radio
and Television Corporation, Aug. 18, 1944.

Div. 4-211. 21-M6

55. Research and Development Conducted by Philco

Corporation on P-4-772 Radio Proximity Fuze for
Large Bombs, Final Report, R. A. Bell, OEMsr-
866, Philco Radio and Television Corporation, June
15, 1943. Div. 4-211. 1-M4

56. [Development of Special Electronic Devices], Re-

port to Division 4y NDRC, on Contract OEMsr-
1003, Final Report, Alan M. Glover and Arnold R.
Moore, OEMsr-1003, Radio Corporation of Amer-
ica, Oct. 23, 1944. Div. 4-231-M3

57. Final Technical Report of Raymond Engineering
Laboratory, Inc., on Work Done under Contract
OEMsr-1378, OEMsr-1378, Report 238, Raymond
Engineering Laboratory, Inc., Oct. 29, 1945.

Div. 4-100-M8

58. [Vacuum Tubes, Types NR-2 (2B-24), NR-3

(2C-27) and NR-5 (2E-27)]. Final Summary Re-
port Regarding Development, A. Abate, OEMsr-
566, Raytheon Manufacturing Company, Oct. 1,
1945. Div. 4-231-M5

59. Contracts OEMsr-1161, OEMsr-1163 BRLG Prox-

imity Fuzes, Final Report, OEMsr-1161 and
OEMsr-1163, Rudolph Wurlitzer Company, Mar.
15, 1945. Div. 4-211. 22-MI

60. Investigation of Rotative Systems of VT-172 and

V T-132 Units, L. M. K. Boelter, University of
California, Department of Engineering, October
1945. Div. 4-232.23-M8

61. Radio Proximity Fuze, Type MROG, OEMsr-749,

Part I, Report WRL-UF-4, University of Florida,
Apr. 2, 1945. Div. 4-211.23-M4

62. Final Chronological Report on Both the RC Proj-

ect and the Mortimer Project, Palmer H. Craig,
OEMsr-749, Report WRL-UF-7, University of
Florida, May 19, 1945. Div. 4-211. 23-M6

63. Summary Technical Report for Contract OEMsr-

769 , OEMsr-769, State University of Iowa, Sept.
29, 1945. Div. 4-100-M7

64. Final Technical Report on Generator Powered
Proximity Fuze for Bombs, Contract II, Western
Electric Company Bell Telephone Laboratory,
May 30, 1944.

65. Photoelectric Fuzes, Final Report, J. F. Wentz,

OEMsr-145 and OEMsr-225, Bell Telephone Labo-
ratories, Mar. 1, 1943. Div. 4-212.2-M4

66. Proximity Fuze, Rocket, Plane-to-Plane, POD
Type, John R. Boykin, OEMsr-343, Termination
Report CFE-761, Westinghouse Electric and
Manufacturing Company, Apr. 28, 1945.

Div. 4-211.1-M5

67. Development of a Ground Approach Proximity
Fuze for Bomb Nose, BRTG, T. M. Bloomer,

452

BIBLIOGRAPHY

OEMsr-343 and OEMsr-1106, Termination Report
CFE-760, Westinghouse Electric and Manufactur-
ing Company, Apr. 28, 1945. Div. 4-211.21-M11

68. Proximity Fuze, Bomb, Nose, Ground Approach:

Type VT, T-82, T. M. Bloomer, OEMsr-343 and
OEMsr-1106, Termination Report CFE-759, West-
inghouse Electric and Manufacturing Company,
Apr. 28, 1945. Div. 4-222.113-M2

69. Proximity Fuze, Hornet, John R. Boykin, OEMsr-

343, Termination Report CFE-762, Westinghouse
Electric and Manufacturing Company, Apr. 28,
1945. Div. 4-211. 1-M6

70. Final Technical Report of Work Performed under

Contract OEMsr-95U, OEMsr-954, Zell Corpora-
tion, Jan. 12, 1945. Div. 4-100-M5

71. Generator Powered Radio Proximity Fuze for

Bombs Transverse Antenna Type, Final Report,
Earl J. Diehl, OEMsr-980 and OEMsr-1133, Serv-
ice Project OD-27, Zenith Radio Corporation, Mar.
30, 1945. Div. 4-211.21-M9

72. Generator Powered Radio Proximity Fuze for
Mortars: Loop Transverse Antenna Type, Earl J.
Diehl, OEMsr-1477, Service Project OD-27, Zenith
Radio Corporation, Oct. 30, 1945.

Div. 4-222. 132-MI

73. Development and Manufacturing Report on NDRC
Gear Reduction Unit for VT Rocket Fuze, OEMsr-
1117, Globe-Union, Inc., Sept. 14, 1945.

Div. 4-238.512-M4

74. RRLG Proximity Fuzes, Final Report, F. H. Os-
borne, OEMsr-1161 and OEMsr-1163, Rudolph
Wurlitzer Company, Mar. 15, 1945.

Div. 4-211. 22-MI

75. Pilot Production of T-50 Fuzes, Allen S. Clarke
and C. N. Julian, OSRD 5351, OEMsr-1227, Serv-
ice Projects OD-27, NO-77B, and NO-77R, Report
A-335, Bowen and Company, Apr. 12, 1945.

Div. 4-222. 111-M3

76. Generator-Powered Proximity Fuzes for Bombs

(Final Technical Report), K. D. Smith and A. L.
Stillwell, OEMsr-905, Bell Telephone Laboratories,
Mar. 24, 1944. Div. 4-211.21-M5

77. [Battery Requirements for Project K-4], Final
Technical Report on Work Performed on Contract
OEMsr-887, C. B. Pear, Jr., OEMsr-887, Wash-
ington Institute of Technology, Feb. 17, 1944.

Div. 4-232.1-M8

78. Generator-Powered Radio Proximity Fuze, Type
T-2005, Muriel E. Pottasch, OEMsr-1437, General
Instrument Corporation, Aug. 1, 1945.

Div. 4-222. 125-M2

Chapter 5

1. Radio Proximity Fuze for Plane-to-Plane Rocket
Application, Harry M. Diamond, W. S. Hinman,
Jr., Robert D. Huntoon, Cledo Brunetti, and C. N.
Page, Service Projects OD-27 and OD-26, Report

A-144, Armor and Ordnance of NDRC, Feb. 12,
1943. Div. 4-211.1-M3

2. The Air Burst Proximity Fuze for Bombs, Rockets,
and Mortars, NBS, Ordnance Development Divi-
sion, National Bureau of Standards, October 1945.

Div. 4-211-M3

3. Computation of Burst Heights of Longitudinally -

Excited Bomb Fuzes, R. P. Schwartz, Report OD-
3-281, NBS, Ordnance Development Division, Aug.
7, 1945. Div. 4-241-M8

4. “VT Rocket Fuzes (for Aircraft Rockets),” Ord-
nance Pamphlet 1470, Apr. 6, 1945.

5. Fuze, Rocket, PD, T-6, TB 9X-93, Dec. 19, 1944.

6. Fuze, Rocket, P.D., T-4 and T-5, TB 9X-94, Dec.
28, 1944.

7. VT Bomb Nose Fuzes, TB 9X-106, Feb. 21, 1945.

8. Test of Fuze, Bomb, Nose T51E1, Army Air
Forces (Eglin Field) S.T.P. 1-45-6, Nov. 27, 1945.

SPECIFICATIONS FOR METAL PARTS
ASSEMBLIES OF FUZES

9. Specifications for the Manufacture and Testing of
the M-3 (MC-382) Radio Fuze, Cledo Brunetti,
NDRC, Division 4, Sept. 30, 1942.

Div. 4-222. 128-MI

10. Technical Specifications for Parts Assemblies for

VT Reaction Grid Detection Fuzes, T-30 and
T-200U, Draft 2, NBS, Ordnance Development
Division, July 20, 1945.* Div. 4-222.126-M2

11. Specification for Longitudinally Excited, Genera-
tor Powered Radio Proximity Fuze, BRLG-100,
NDRC, Division 4, Feb. 25, 1944.

Div. 4-211.21-M4

12. Fuze, Bomb, Nose, VT, M-168, Parts Assembly,
Tentative Specification, Ordnance Department,
U. S. Army, AXS-1691, Apr. 18, 1946.

13. Specification for Transversely Excited, Generator

Powered Radio Proximity Fuze, T-51E1, NDRC,
Division 4, Jan. 5, 1945. Div. 4-211. 21-M8

14. Fuze, Bomb, Nose, VT, T-82E2, Parts Assembly,
Tentative Specification, Ordnance Department,
U. S. Army, AXS-1610, July 19, 1945.

15. Fuze, VT, T-132, Parts Assembly, Tentative Speci-
fication, Ordnance Department, U. S. Army, AXS-
1615, July 1, 1945.

16. Fuze, VT, T-171, Parts Assembly, Tentative Speci-
fication, Ordnance Department, U. S. Army, AXS-
1667, July 23, 1945.

Chapter 6

ARMOR AND ORDNANCE REPORTS OF NDRC

1. Generator-Powered Radio Proximity Fuze for
Bombs Transverse Antqypia Type, Earl J. Diehl,
OSRD 5111, OEMsr-980 and OEMsr-1133, Service

* No official specification was published for the OD
models. The specifications for the OD and RGD models
are quite similar except for the RF loading procedure
and except for the audio input test circuit.

BIBLIOGRAPHY

453

Projects OD-27 and NO-77B, Final Report A-326,
Zenith Radio Corporation, Mar. 30, 1945.

Div. 4-211.21-M10

2. Pilot Production of T-50 Fuzes, Allen S. Clarke
and C. N. Julian, OSRD 5351, OEMsr-1227, Serv-
ice Projects OD-27, NO-77B, and NO-77R, Report
A-335, Bowen and Company, Apr. 12, 1945.

Div. 4-222.111-M3

REPORTS OF ORDNANCE DEVELOPMENT DIVI-
SION OF NATIONAL BUREAU OF STANDARDS

3. Engineering Letters Nos. 1 to 63, Inclusive, Cover-

ing tine Period May 27, 19 UU to August 29, 1945
(No. 58 not microfilmed), NBS, Ordnance De-
velopment Division. Div. 4-100-M3

(The aforementioned Engineering Letters cover
a variety of items relating to production problems
for radio proximity fuzes. They were prepared by
the Production Engineering Section of the Ord-
nance Development Division and transmitted to
the various manufacturers engaged in production
of fuzes.)

REPORTS OF CONTRACTORS OF DIVISION U
OF NDRC

4. Summary of Activities in Development and Pilot

Manufacturing Run of Radio Fuzes and Acces-
sories, Vernon D. Hauck, OEMsr-258, Friez In-
strument Division, Bendix Aviation Corporation,
Sept. 27, 1944. Div. 4-100-M4

5. Preliminary Draft of Final Technical Report
under Contracts OEMsr-885 and OEMsr-1113,
OEMsr-885 and OEMsr-1113, Emerson Radio and
Phonograph Corporation, May 14, 1945.

Div. 4-100-M6

6. [Development of the 7-mm Rectifier Disc], Final

Report of the Federal Telephone and Radio Cor-
poration, T. Smith Taylor, OEMsr-941, Oct. 5,
1943. Div. 4-235-M4

7. Generator-Powered Radio Proximity Fuze, Type
T-2005, Muriel E. Pottasch, OEMsr-1437, General
Instrument Corporation, Aug. 1, 1945.

Div. 4-222. 125-M2

8. Generator-Powered Radio Proximity Fuze for

Mortars, Longitudinal Excitation Type, T-132,
Alfred S. Khouri, OEMsr-1117, Globe-Union, Inc.,
Sept. 30, 1945. Div. 4-222.131-M6

9. [Alnico Rotor Generators], Final Report, Con-
tract OEMsr-1134, C. W. Clemons, OEMsr-1134,
Knapp-Monarch Company, Nov. 20, 1944.

Div. 4-232. 2-M19

10. [Generators], Final Report on Contract OEMsr-

981, C. W. Clemons, OEMsr-981, Knapp-Monarch
Company, Feb. 17, 1944. Div. 4-232.2-M11

11. Final Progress Report, Contract OEMsr-1196,

Maurice E. Swift, OEMsr-1196, Philco Corpora-
tion, May 31, 1945. Div. 4-211.21-M12

12. Final Technical Report on Generator-Powered

Proximity Fuzes for Bombs, K. D. Smith and A. L.
Stillwell, OEMsr-905, Contract II, Western Elec-
tric Company, Bell Telephone Laboratories, May
30, 1944. Div. 4-211.21-M5

13. Development of a Ground Approach Proximity
Fuze for Bomb, Nose, BRTG, T. M. Bloomer,
OEMsr-343 and OEMsr-1106, Termination Report
CFE-760, Westinghouse Electric and Manufac-
turing Company, Apr. 28, 1945.

Div. 4-211. 21-M11

14. Proximity Fuze, Bomb, Nose, Ground Approach,

Type VT, T-82, T. M. Bloomer, OEMsr-343 and
OEMsr-1106, Termination Report CFE-757, West-
inghouse Electric and Manufacturing Company,
Apr. 28, 1945. Div. 4-222.113-M2

15. Final Technical Report of Work Performed under

Contract OEMsr-954, OEMsr-954, Zell Corpora-
tion, Jan. 12, 1945. Div. 4-100-M5

16. Generator-Powered Radio Proximity Fuze for
Mortars, Loop Transverse-Antenna Type, Earl J.
Diehl, OEMsr-1477, Service Project OD-27, Zenith
Radio Corporation, Oct. 30, 1945.

Div. 4-222. 132-MI

17. Mass Production of T-51 Fuzes by Zenith Radio
Corporation (OEMsr-980 and OEMsr-1133), Oct.
3, 1945.

Chapter 7

ARMOR AND ORDNANCE REPORTS OF NDRC

1. Analysis of Feedback Amplifiers for MC-382

Fuzes. Robert D. Huntoon, William L. Kraushaar,
and Herbert D. Cook, Progress Report A122, Dec.
7, 1942. Div. 4-238. 222-MI

2. Pilot Production of T-50 Fuzes, Allen S. Clarke
and C. N. Julian, OSRD 5351, OEMsr-1227, Serv-
ice Projects OD-27, NO-77B and NO-77R, Report
A-335, Bowen and Company, Apr. 12, 1945.

Div. 4-222. 111-M3

NDRC REPORTS AND MEMORANDA

3. Radiation Properties of BRLG, Robert D. Hun-

toon, Service Project OD-27, Memorandum Report
43-R, July 29, 1942. Div. 4-243.11-MI

4. Description of 1000-G Centrifuge, Allen S. Clarke,
Eng. Memorandum, Nov. 25, 1942. Div. 4-615-MI

5. Engineering Report on MC-382 Test Equipment,
Preliminary Draft, Nov. 26, 1942.

Div. 4-222.128-M3

NDRS SPECIFICATIONS

6. Specification for Electron Tube NR-2A, a Diode

Tube, Aug. 1, 1944. Div. 4-231.2-M3

7. ' Specification for Generator G-l, NBS, Ordnance

Development Division, Nov. 25, 1944.

Div. 4-232.2-M20

454

BIBLIOGRAPHY

REPORTS OF ORDNANCE DEVELOPMENT DIVI-
SION OF NATIONAL BUREAU OF STANDARDS

8. Electronic Frequency Meter , Charles Ravitsky,
Leonard C. Pochop, and J. G. Reid, Jr., Report 0D-

2- 15 (First Series), NBS, Ordnance Development

Division, July 26, 1943. Div. 4-613-MI

9. Rotary Shaker for Pre-Testing BRLG Heads ,
Robert D. Huntoon, Report OD-3-7, NBS, Ord-
nance Development Division, Oct. 22, 1943.

Div. 4-614-MI

10. Critical Grid Voltage of Thyratrons and Hum

Voltage Output of BRLG-11, F. Lamar Cooke,
Report OD-3-9, NBS, Ordnance Development Divi-
sion, Oct. 27, 1943. Div. 4-231.1-M7

11. Methods of Measuring the Critical Voltage of

Thyratrons, F. Lamar Cooke, Report OD-3-13,
NBS, Ordnance Development Division, Oct. 28,
1943. Revised: Nov. 9, 1943. Div. 4-231. 1-M8

12. Generator Performance, William L. Kraushaar,

Report OD-3-17, NBS, Ordnance Development Di-
vision, Nov. 1, 1943. Div. 4-232.2-M8

13. Preliminary Report on Tuning and Loading Device
for BRLG, Paul E. Landis, Report OD-3-37, NBS,
Ordnance Development Division, Nov. 29, 1943.

Div. 4-233. 1-MI

14. Tuning BRLG, Robert D. Huntoon, Report OD-

3- 87, NBS, Ordnance Development Division, Jan.

29, 1944. Div. 4-233.1-M2

15. BRLG Tuning on Various Vehicles, Bertrand J.

Miller, Report OD-3-106 and Addendum OD-3-
106A, Mar. 3 and 20, 1944. Div. 4-231.2-M2

16. Addendum to Report OD-3-106, Bertrand J. Miller
and Charles C. Gordon, Report OD-3-106A, NBS,
Ordnance Development Division, Mar. 20, 1944.

Div. 4-231.2-M2

17. Microphonic Stability of Oscillator-Diode Type

Fuze Circuits, Robert D. Huntoon, Report OD-3-
117, NBS, Ordnance Development Division, Mar.
22, 1944. Div. 4-238.31-MI

18. Loading Circuit for Final Test Chamber to Be
Used at W Frequency ayid Encasing Cup Speci-
fication, Thomas C. Bagg, Report OD-3-126, NBS,
Ordnance Development Division, Apr. 1, 1944.

Div. 4-233.1-M3

19. Testing RGD Units, Philip Krupen, Report OD-

3-131, NBS, Ordnance Development Division, Apr.
22, 1944. Div. 4-238.32-M6

20. Dummy Antennas, Robert D. Huntoon, Report OD-

3-133, NBS, Ordnance Development Division, Apr,
29, 1944. Div. 4-233-M3

21. Preliminary Investigation of Characteristics of

Test Chamber, with Respect to Relative Position
of Unit Therein, J. L. Pike and Otto E. Spokas,
Report OD-3-135, NBS, Ordnance Development
Division, Apr. 25, 1944. Div. 4-233. 1-M4

22. Electronic Tachometer, Herbert D. Cook, Report

OD-3-137, NBS, Ordnance Development Division,
July 28, 1944. Div. 4-621.1-MI

23. Triode Microphonics, Robert D. Huntoon, Report

OD-3-153, NBS, Ordnance Development Division,
May 20, 1944. Div. 4-231.3-M2

24. Compensated Resistors for Tuning and Loading

Standards, E. Eisner and Paul T. Hawes, Report
OD-3-154, NBS, Ordnance Development Division,
May 24, 1944. Div. 4-236-M4

25. Resonant Loading of BRTG Units by Test Boxes,
Ralph Stair, Glenn L. Scillian, and Leonard C.
Pochop, Report OD-3-196, NBS, Ordnance De-
velopment Division, Nov. 13, 1944.

Div. 4-233.1-M7

26. The T-132 (Mortar Fuze) Apex Performance

Problem, William L. Kraushaar, Report OD-3-220,
NBS, Ordnance Development Division, Mar. 3,
1945. Div. 4-222. 131-M2

27. Test Fixture for Balancing the Single Bearing

Nose Assembly, Jacob Rabinow, Supplement 3 to
Report OD-4-48, NBS, Ordnance Development
Division, Jan. 13, 1945. Div. 4-616-M6

28. Compression Test Equipment, C. Chartock, Report

OD-4-50, NBS, Ordnance Development Division,
Apr. 27, 1944. Div. 4-623-MI

29. Torsion Wire Dynamometer, Louis Schuman, Re-

port OD-4-105, NBS, Ordnance Development Divi-
sion, May 26, 1945. Div. 4-612-M2

30. Proposed Design for Dynamic Balancing Machine,
Jacob Rabinow, Report OD-4-108, NBS, Ordnance
Development Division, June 6, 1945.

Div. 4-616-M8

31. Report of Shelf Life Test on MC-382 Unit, Paul J.
Martin, Report OD-5-522, NBS, Ordnance De-
velopment Division, Oct. 12, 1944.

Div. 4-238.222-M5

32. A Study of the Development of the BRLG-100

Specifications of February 25, 19 UU, Report OD-
5-617, NBS, Ordnance Development Division, Sept.
1, 1944. Div. 4-211.21-M7

33. A Study of the Development of the Specification

for the Rectifier Bridge Assembly RA-1 of July
5, 19 Uh, Report OD-5-637, NBS, Ordnance Develop-
ment Division, Oct. 4, 1944. Div. 4-235-M2

34. A Study of the Development of the NDRC Speci-
fication for Generator G-l Dated February 25,
19 UU, Report OD-5-645, NBS, Ordnance Develop-
ment Division, Oct. 6, 1944. Div. 4-232. 2-M 17

35. A Study of the Development of the Specifications

for NR-2A Diode, NR-3 /NS-3 Triode, NS-A Thy-
ratron and NR-5 / NS-5 Pentode, Dated August 1,
19 H, Report OD-5-671, NBS, Ordnance Develop-
ment Division, Oct. 20, 1944. Div. 4-231-M2

36. Zenith Revised Final Test Position, Paul E.

Landis, Report OD-5-787, NBS, Ordnance Develop-
ment Division, Apr. 16, 1945. Div. 4-622-M3

37. Mechanical Properties of Final Test Chamber,

Robert D. Huntoon and T. F. Protz, Report OD-
BE-9R, NBS, Ordnance Development Division,
July 24, 1944. Div. 4-619-M2

38. A New Proposal for Shaking Each Unit in Final

CHET

BIBLIOGRAPHY

455

Test, Wendell L. Lees, Report OD-BEr72R, NBS,
Ordnance Development Division, Feb. 24, 1945.

Div. 4-622-M2

39. Compensated Tuning Resistors Used in Tuning

T-30 Fuzes for Aircraft Rockets (AR and HVAR),
Paul T. Hawes and Thomas C. Bagg, Report OD-
TEG-6R, NBS, Ordnance Development Division,
Dec. 14, 1944. Div. 4-236-M7

40. Test Line for T-132 Unit, Globe-Union and Wur-

litzer Model, Thomas C. Bagg, Engineering Report
OD-2-TEG-SR, NBS, Ordnance Development Di-
vision, Jan. 30, 1945. Div. 4-222.131-MI

41. Measurement of Firing Voltage, Robert D. Hun-

toon, Project OD-3, NBS, Ordnance Development
Division, Aug. 20, 1943. Div. 4-621-M2

42. Electronic Demagnetizer, Engineering Letter 40,
Jan. 3, 1945.

MEMORANDA OF ORDNANCE DEVELOPMENT
DIVISION OF NATIONAL BUREAU
OF STANDARDS

43. Loading Device for BRTG Units, L. A. Riley and
G. J. Tedore, Memorandum OD-5-88M, NBS, Ord-
nance Development Division, Dec. 26, 1944.

Div. 4-233.1-M8

44. Compensated Versus Uncompensated Resistors
for Sensitivity Measurements on RGD Units, Paul
E. Landis, Memorandum OD-5-242M, NBS, Ord-
nance Development Division, June 25, 1945.

Div. 4-328.2-M2

45. Noise Differences in Final Test Chambers, Robert
D. Huntoon, Memorandum OD-BE-11M, NBS,
Ordnance Development Division, June 26, 1944.

Div. 4-233.1-M6

46. Blocking Voltage for Use in Making Audio Test

on OD Units, Herbert D. Cook, Memorandum OD-
TEG-35M, NBS, Ordnance Development Division,
Feb. 12, 1945. Div. 4-621-M4

47. Hum Injection Adjustment, Charles R. Duke,

Herbert D. Cook, and Thomas C. Bagg, Memo-
randum OD-TEG-78M, NBS, Ordnance Develop-
ment Division, July 28, 1945. Div. 4-621-M5

48. Tuning and Adjustment of MC-382. Memorandum
to Harry Diamond from W. S. Hinman, Jr., NBS,
Ordnance Development Division, Nov. 16, 1942.

Div. 4-222.128-M2

REPORTS OF CONTRACTORS OF
DIVISION U NDRC

49. Radiation Dummy Load Consideration, MC-382,

R. H. Pintell, Service Project OD-27, Memorandum

Report 33R, Emerson Radio and Phonograph Cor-
poration, Mar. 2, 1943. Div. 4-243.12-MI

50. Development of Balancing Equipment for T-171
Turbine Rotor Assemblies, M. S. Redden and Allen

S. Clarke, OEMsr-1227, Bowen and Company,

May 1945. Div. 4-616-M7

51. Mass Production of T-51 Fuzes by Zenith Radio

Corporation, Earl J. Diehl, OEMsr-1477, Service
Project OD-27, Zenith Report of Contract W-28-
004-SC-965, Oct. 30, 1945. Div. 4-222.112-M5

U. S. MILITARY REPORTS

52. Tentative Specifications for Rectifiers, AXS-1613,
Mar. 31, 1945.

53. Tentative Specifications for Tubes, Vacuum, and
Gas Filled. Ordnance Department, U. S. Army,
AXS-1612 (Revision 1), July 25, 1945.

54. War Department Technical Manual for Quality
Control Testing for Ring-Type and Bar-Type VT
Nose Metal Parts Assemblies for Bombs and
Rockets.

UNCLASSIFIED TECHNICAL PUBLICATIONS

55. Measurements of Admittance at UHF. J. M. Miller
and B. Salzberg, RCA Review 3, April 39, p. 486.

DRAWING REFERENCES

Drawing Reference
N umber
1
2

8
9

NBS Drawing Index
L5515
L5516
L5524
L5526
L5529
L5530
L5531
L5532
L5533

Chapter 8

ARMOR AND ORDNANCE REPORTS AND
MEMORANDA OF NDRC

1. Radio Reporters for Proximity Fuze Testing,

Allen V. Astin, OSRD 589, Report A-53, May 21,
1942. Div. 4-611-MI

2. Proving Ground Operations and Facilities for

Testing Proximity Fuzes for Bombs and Rockets,
Lauriston S. Taylor, OSRD 719, Memorandum
A-44M, July 20, 1942. Div. 4-222.129-MI

3. Note on a Practical Method for the Field Testing
of Radio Proximity Fuzes for Rocket Applications,
Harry M. Diamond and W. S. Hinman, Jr., OSRD
767, Memorandum A-48M, July 30, 1942.

Div. 4-222. 129-M2

4. Sampling Formulas for Qualification and Proof
Testing of Production Lots, T. M. White, OSRD
3198, Memorandum A-82M, January 1944.

Div. 4-770-MI

NDRC ENGINEERING REPORTS

5. Proposed Proof Range for M-2 and M-3 Fuzes at
Aberdeen, Harry M. Diamond, Service Project
OD-27, Memorandum Report 1-M, Dec. 29, 1942.

Div. 4-222. 223-MI

456

BIBLIOGRAPHY

6. Chapter 4 (“Exterior Ballistics”) of Rocket Fun-

damentals ; prepared under the auspices of Section
H, Division 3, edited by Bryce L. Crawford, Jr.,
OSRD 3992, OEMsr-273, Report ABL-SR4, George
Washington University, 1944. Div. 4-410-MI

7. Chapter 13 (“Flight Tests of Rockets”) of Rocket

Fundamentals, prepared under the auspices of
Section H, Division 3 (OSRD 3992), Report ABL-
SR4, 1944. Div. 4-410-MI

REPORTS AND MEMORANDA OF ORDNANCE
DEVELOPMENT DIVISION OF NATIONAL
BUREAU OF STANDARDS

8. Frequency of Yaw of Budd UVz" Rockets Fired
from a Plane, Theodore B. Godfrey, Service Proj-
ect OD-27, Memorandum Report 47-T, NBS, Ord-
nance Development Division, Feb. 11, 1943.

Div. 4-412.2-MI

9. Three-Dimensional Analysis of 11 Trajectories of
PEP-M2 Fuzes Fired from a Plane at Aberdeen,
January 23 and 2U, 19 U3, Theodore B. Godfrey,
Service Project OD-27, Memorandum Report 54-T,
NBS, Ordnance Development Division, Feb. 11,

1943. Div. 4-222-224-M5

10. Yaw Reporter Test, Theodore B. Godfrey and

L. C. Miller, Service Project OD-27, Memorandum
Report 401-T, NBS, Ordnance Development Divi-
sion, Aug. 9, 1943. Div. 4-412.2-M2

11. Salvo Firing in Search of Sympathetic Function-
ing of MC-380 and MC-382 Fuzes, F. R. Kotter
and T. N. White, Report OD-1-15, NBS, Ordnance
Development Division, Sept. 23, 1943.

Div. 4-245-MI

12. Puff Delay, 500-lb Bomb, Theodore B. Godfrey,

Report OD-1-41, NBS, Ordnance Development
Division, Nov. 5, 1943. Div. 4-242.13-MI

13. A Modified Method of Scanning Phonograms, J. J.
Hopfield, Report OD-1-130, NBS, Ordnance De-
velopment Division, Feb. 5, 1944. Div. 4-617-MI

14. Field Test of SW200 0.7-Sec Switches; Photo-
graphic Method for Timing Early Functions in
High Angle Firing, H. F. Stimson, R. G. Tobey,
and D. W. Scott, Report OD-1-237, NBS, Ordnance
Development Division, Apr. 20, 1944.

Div. 4-238. 511-M5

15. Static Tests of BRLG Function Indicators, T. C.

Hellmers, L. L. Parker, and L. C. Miller, Report
OD-1-272, NBS, Ordnance Development Division,
May 3, 1944. Div. 4-626-MI

16. UO Bowen T-50 E10 on Refrigerated Mk 7, D. A.
Worcester and D. W. Scott, Report OD-1-529,
NBS, Ordnance Development Division, Oct. 20,

1944. Div. 4-222. 127-MI

17. Field Test, Rotation of M9A1 with Hand-Crimped

Fins, R. G. Tobey and D. W. Scott, Report OD-
1-588, NBS, Ordnance Development Division, Dec.
18, 1944. Div. 4-412.2-M3

18. Static Tests to Determine the Effect of Different
Trap and Motor Combinations on the Functioning

of the T-5 Fuze, H. F. Stimson, John Beek, Jr.,
E. Allen Cook, and Charles C. Gordon, Report
OD-1-589, NBS, Ordnance Development Division,
Dec. 15, 1944. Div. 4-222.121-M7

19. Ballistics of Mk 1 and Mk 7 Motors with T-50

and T-51 Units and Slip Factor Data for Various
Vehicles, D. C. Friedman and G. L. Rabinow, Re-
port OD-1-591, NBS, Ordnance Development Divi-
sion, Dec. 21, 1944. Div. 4-411. 1-M5

20. Plane Firing of T-30 and Mk 7, D. W. Scott, Re-

port OD-1-650, NBS, Ordnance Development Divi-
sion, Feb. 7, 1945. Div. 4-222.124-MI

21. Effect of Rocket Spin upon the Performance of
VT Fuzes T-U, T-5, T-6, Theodore B. Godfrey,
Summary Report OD-1-668, NBS, Ordnance De-
velopment Division, Mar. 13, 1945.

Div. 4-222. 123-M3

22. Plane Firing, Philco T-200U on T-87 , D. A. Wor-
cester and D. W. Scott, Report OD-1-744, NBS,
Ordnance Development Division, May 10, 1945.

Div. 4-222. 125-Ml

23. Visibility of Various Mortar Spotting Charges,

R. G. Tobey and G. Rabinow, Report OD-1-829,
NBS, Ordnance Development Division, July 11,

1945. Div. 4-626-M2

24. Afterburning from Rocket Motors and Malfunc-

tioning of VT Fuzes, H. F. Stimson, Report OD-
1-896, NBS, Ordnance Development Division, Oct.
15, 1945. Div. 4-411. 11-M6

25. An Investigation of Mo7'tar-Shell Muzzle Veloc-
ities, H. V. Menapace, M. H. Seibel, and G. L.
Rabinow, Report OD-1-909, NBS, Ordnance De-
velopment Division, Mar. 14, 1946. Div. 4-515-MI

26. A Method of Recording Size and Concentration of
Raindrops, Theodore B. Godfrey, R. K. Pickels,
and D. A. Worcester, Report OD-1-920, NBS, Ord-
nance Development Division, May 21, 1946.

Div. 4-740-MI

27. Equivalent Release Co7iditio7is for Level Flight

Bombmg and Dive Bombing, Irene Freuder, F. L.
Celauro, and T. N. White, Technical Memorandum
OD-1-TM2, NBS, Ordnance Development Divi-
sion, Oct. 30, 1945. Div. 4-211. 3-M5

28. Audio Limiter, W. A. Yates, Technical Memoran-

dum OD-1-TM5, NBS, Ordnance Development Di-
vision, Oct. 29, 1945. Div. 4-617-M2

29. Recording Oscilloscope and 16-mm Eastman
Oscilloscope Camera, N. Newman, Technical
Memorandum OD-1-TM8, NBS, Ordnance De-
velopment Division, Nov. 2, 1945.

Div. 4-617-M3

30. Intermittent Reco7'ding Control, N. Newman,
Technical Memorandum OD-1-TM9, NBS, Ord-
nance Development Division, Nov. 7, 1945.

Div. 4-617-M4

31. Fifty-Cycle Oscillator, N. Newman, Technical
Memorandum OD-1-TM10, NBS, Ordnance De-
velopment Division, Nov. 7, 1945. Div. 4-619-M5

32. Notes on Loading, Assembly a7id Storage Pro-

BIBLIOGRAPHY

457

cedures in Rocket Testing at Blossom Point Prov-
ing Ground, R. G. Robey and L. T. Johnson, Tech-
nical Memorandum OD-1-TM19, NBS, Ordnance
Development Division, Sept. 25, 1945.

Div. 4-412.4-M7

33. Notes on Mock-Plane Target, Rocket Launchers
and Firing Procedures at Blossom Point, A. P.
Sutten, Technical Memorandum OD-1-TM20, NBS,
Ordnance Development Division, Sept. 25, 1945.

Div. 4-412. 4-M8

34. Notes on Drainage, Firing Tower Construction,
Fire Prevention and Observational Procedures at
Blossom Point Proving Ground, R. G. Tobey, Tech-
nical Memorandum OD-1-TM21, NBS, Ordnance
Development Division Sept. 25, 1945.

Div. 4-412.4-M9

35. Navy Rocket Trajectory Analysis, A. L. Leiner,

Memorandum OD-2-203, NBS, Ordnance Develop-
ment Division, May 5, 1945. Div. 4-412.1-M8

36. Standard Statistical Methods for Testing the
Difference between Mean Values, B. M. Bennett,
Memorandum OD-2-205M, NBS, Ordnance De-
velopment Division, May 7, 1945. Div. 4-770-M3

37. Early Functions of MC-382 Radio-operated Plane-

to-Plane Rocket Fuze, Bertrand J. Miller and
Robert D. Huntoon, Progress Report OD-3-AB2,
NBS, Ordnance Development Division, June 8,
1943. Div. 4-222. 128-M12

38. Tests BJM-5 and BJM-6, Charles Ravitsky, Prog-
ress Report OD-7-206R, NBS, Ordnance Develop-
ment Division, May 14, 1945. Div. 4-222. 124-M3

39. Sequential Analysis of Statistical Data: Applica-

tions, T. N. White and H. C. Doob, Report OD-
OAG-46, NBS, Ordnance Development Division,
Sept. 27, 1944. Div. 4-770-M2

40. [Cenco Rocket] Motor, NBS Drawing 440 R, NBS,
Ordnance Development Division, May 20, 1942.

Div. 4-411.1-M2

REPORTS OF CONTRACTORS OF DIVISION U
OF NDRC

41. Chapter VI of Summary Technical Report, Con-
tract OEMsr-769, submitted by James A. Jacobs,
State University of Iowa, Sept. 29, 1945.

Div. 4-100-M7

42. Chapter VII of Summary Technical Report, Con-
tract OEMsr-769, submitted by James A. Jacobs,
State University of Iowa, Sept. 29, 1945.

Div. 4-100-M7

43. Clinton Field Station Report 60, State University
of Iowa.

44. Mortar Fuze Recovery, W. E. Nickell OEMsr-769,

Report MB-3-1-45, State University of Iowa, Mar.
31, 1945. Div. 4-619-M4

45. The Calculation of Trajectories, L. E. Ward,
OEMsr-769, Technical Report T3-8-1-45, State
University of Iowa, Aug. 29, 1945. Div. 4-512-M3

46. The Effects on Trajectories of Small Changes in
Initial Conditions with Aioplication to Wind Cor-

rections, L. E. Ward, OEMsr-769, Technical Re-
port T3-9-1-45, State University of Iowa, Sept. 12,
1945. Div. 4-512-M4

UNITED STATES MILITARY PUBLICATIONS

47. War Department Manual TM11-2U10.

48. Handbook of Instructions and Parts Catalog
AN 10-25-50.

Chapter 9

ARMOR AND ORDNANCE REPORTS OF NDRC

1. Radio Proximity Fuze for Plane-to-Plane Rocket
Application, Harry M. Diamond, W. S. Hinman,
Jr., Robert D. Huntoon, Cledo Brunetti, and
Chester H. Page, Service Projects OD-27 and
OD-26, Report A-144, Feb. 12, 1943.

Div. 4-211.1-M3

2. Sampling Formulas for Qualifications and Proof
Testing of Production Lots, T. N. White, OSRD
3198, Memorandum A-82M, January 1944.

Div. 4-770-MI

3. Reports Pertinent to Early and Middle Function-
ing of MC-382 Fuze, as follows:

3a. A Study of the Relation between Afterburn-
ing and Thyratron Voltage, R. Vorkink, Serv-
ice Project OD-27, Memorandum Report
158-T, Apr. 14, 1943. Div. 4-238.212-M3

3b. Tests with Eccentric and with Non-Eccentric
Powder, High-Angle Firing, R. Vorkink,
Service Project OD-27, Memorandum Report
338-T, June 1943. Div. 4-222.128-M11

3c. Test for Ride Through with Various Powders
and Firing Angles, R. Vorkink, Service
Project OD-27, Memorandum Report 383-T,
Aug. 5, 1943. Div. 4-222.128-M14

3d. Fuze T6, Range, Dispersion, and Water Ap-
proach Function, D. C. Friedman, Service
Project OD-27, Memorandum Report 388-T,
July 28, 1943. Div. 4-222.128-M13

3e. Test of Effect of Velocity on Early Function-
ing, R. Vorkink, Service Project OD-27,
Memorandum Report 405-T, Aug. 12, 1943.

Div. 4-222. 128-M15

4. A Comparison of Several Makes of MC-382 Fuze
with Respect to Early, Target and Late Functions
and Duds, T. N. White, Memorandum Report
220-T, May 13, 1943, and Supplement to A Com-
parison of Several Makes of MC-382 Fuze with
Respect to Early, Target and Late Functions and
Duds, T. N. White, Service Project OD-27, Mem-
orandum Report 282-T, June 10, 1943.

Div. 4-222.128-M10

REPORTS OF ORDNANCE DEVELOPMENT
DIVISION OF NATIONAL BUREAU
OF STANDARDS

5. Reports pertinent to early and middle function-
ing of the MC-382, as follows:

458

BIBLIOGRAPHY

5a. Relation between Early Function and After-
burning, T. N. White, Report OD-1-AB1,
NBS, Ordnance Development Division, Mar.

17, 1943. Div. 4-222. 129-M3

5b. Effect of Powder Lot on Afterburning and

Slivers, L. C. Miller, Report OD-1-AB2,
NBS, Ordnance Development Division, Mar.

18, 1943. Div. 4-411. 11-M2

5c. The Effect of Powder Load on Afterburning

and Slivers, L. C. Miller, Report OD-1-AB3,
NBS, Ordnance Development Division, Mar.
20, 1943. Div. 4-222.128-M4

5d. Effect of Fin Structure on Early Function-
ing, L. C. Miller, Report OD-1-AB4, NBS,
Ordnance Development Division, Mar. 23,
1943. Div. 4-222. 128-M5

5e. Early Function Tests (1) Fuzes with Re-
duced Sensitivity, (2) Motors with Metal
Sweeps, L. C. Miller, Report OD-1-AB5,
NBS, Ordnance Development Division, Mar.
23, 1943. Div. 4-222.128-M6

5f. Early Functions with MC-382 Fuze, Further
Testing with Sweeps and with Powders,
T. N. White, Report OD-1-AB6, NBS, Ord-
nance Development Division, Mar. 27, 1943.

Div. 4-222. 128-M7
5g. Experiments on Early Functioning with
Revere Motors (1) Soldering of Fin Retain-
ing Rings, (2) Test of Powder Lot 9978, (3)
Soldering of Fins in Open Position, L. C.
Miller, Report OD-1-AB7, NBS, Ordnance
Development Division, Mar. 31, 1943.

Div. 4-411.2-M2

5h. Static Tests on Afterburning (1) Use of
Metal Sweeps, (2) Use of J P-26 5 Powder,
L. C. Miller, Preliminary Report OD-1-AB8,
NBS, Ordnance Development Division, Mar.
29, 1943. Div. 4-411. 11-M3

5i. Progress Report on Afterburning , H. F.
Stimson, Progress Report OD-1-AB9, NBS,
Ordnance Development Division, Apr. 9,
1943. Div. 4-411. 11-M4

5j. High-Angle Firing with MC-382 Fuzes
[ Part ] A. Early Function Tests (1) Detun-
ing of Units (2) Use of Sweeps and Plugs;
[ Part] B. Tests of Mechanical SD Switches,
L. C. Miller, Final Report OD-1-AB11, NBS,
Ordnance Development Division, Apr. 17,
1943. Div. 4-222. 128-M8

5k. Incidence of Early Functions with POD
Type Fuzes and MC-382 Fuzes: Compari-
sons Based on Target Function and High-
Angle Firing Tests, T. N. White, Report
OD-1-AB12, NBS, Ordnance Development
Division, May 1, 1943. Div. 4-222.129-M4
51. Tests of Sweeps and Plugs, R. Vorkink, Re-
port OD-1-AB13, NBS, Ordnance Develop-
ment Division, May 7, 1943.

Div. 4-222.128-M9

5m. High-Angle Night Firing with Powders
A-20, A-21, and A-22: Afterburning , Burn-
ing Distances, H. F. Stimson, Report OD-1-
AB14, NBS, Ordnance Development Di-

vision, May 13, 1943. Div. 4-411. 11-M5
5n. RC Delay Added to SW-200 Arming

Switches, Effect on Early Functioning of
MC-382 Fuzes, T. N. White, Report OD-1-
AB15, NBS, Ordnance Development Di-

vision, Sept. 14, 1943. Div. 4-222.128-18
5o. Tests for Early Functioning with Different
Powder Weights, R. Vorkink, Report OD-1-
AB16, NBS, Ordnance Development Di-

vision, Aug. 26, 1943. Div. 4-222.128-M16
5p. Test for Mal-Functions of MC-382 with
Special Fin Motors (No Locking Burr), R.
Vorkink, Report OD-1-1, NBS, Ordnance
Development Division, Sept. 2, 1943.

Div. 4-222. 128-M17
5q. Tests on Early Functioning of MC-382
Fuzes [Part] A. Use of Purge Pellets;
[ Part ] B. Increased Surface Area of Pro-
pellant, L. C. Miller, Report OD-1-5, NBS,
Ordnance Development Division, Sept. 14,
1943. Div. 4-222. 128-M19

5r. Effect of Propellant on Early Functioning
[Part] A. Amount of Regular Propellant ;
[Part] B. Special Propellant; [Part] C.
Purge Pellets, T. N. White, Report OD-1-8,
NBS, Ordnance Development Division, Sept.
21, 1943. Div. 4-222. 128-M21

5s. Test of Propellant Charge on Early Func-
tioning, R. Vorkink, Report OD-1-13, NBS,
Ordnance Development Division, Sept. 20,
1943. Div. 4-222. 128-M20

5t. Test of Effect of Purge Pellets on Early
Functioning, R. Vorkink, Report OD-1-17,
NBS, Ordnance Development Division, Sept.
30, 1943. Div. 4-222.128-M22

5u. Early Functioning of MC-382 Fuzes, Purge
Pellet Field Test 5, L. C. Miller, Report OD-
1-22, NBS, Ordnance Development Division,
Oct. 6, 1943. Div. 4-222.128-M23

5v. Early Functioning of MC-382 Fuzes, Purge
Pellet Field Tests 6 and 7, L. C. Miller, Re-
port OD-1-24, NBS, Ordnance Development
Division, Oct. 13, 1943. Div. 4-222.128-M24
5w. MC-382 Fuze Performance as Affected by

Motors with Non-Locking Type Fins, T. N.
White, L. C. Miller, and R. Vorkink, Report
OD-1-27, NBS, Ordnance Development Di-
vision, Oct. 15, 1943. Div. 4-222.128-M25
5x. MC-382 Fuze Performance as Affected by

Motors with Fins Welded into the Opened
Position, D. C. Friedman, Report OD-1-40,
NBS, Ordnance Development Division, Nov.
4, 1943. Div. 4-222.128-M26

5y. Early Functioning of MC-382 Fuze, Purge
Pellet Field Test 8 (Also Tests with POD

BIBLIOGRAPHY

459

Type Fuzes and with Pressure-Control
Valves ), T. N. White and R. Vorkink, Re-
port OD-1-42, NBS, Ordnance Development
Division, Nov. 19, 1943. Div. 4-222.128-M27
5z. Purge Pellet Test 9 Including Test of (1)
Combination of Motors and Propellants, (2)

A New Salted Powder, (3) Pressure-Control
Valves, R. Vorkink, Report OD-1-59, NBS,
Ordnance Deevlopment Division, Nov. 23,

1943. Div. 4-222. 128-M28

5aa. High-Angle Test of MC-382 Units to Deter-
mine Propellant-Motor Combination for
Acceptance Testing, D. C. Friedman, Report
OD-1-119, NBS, Ordnance Development Di-
vision, Jan. 26, 1944. Div. 4-222.128-M29

5bb. Field Test of Eight Lots of Pellets, R. Vor-
kink, Report OD-1-125, NBS, Ordnance De-
velopment Division, Jan. 29, 1944.

Div. 4-222. 128-M30
5cc. Test of T 5 and T6 on Motors with Spring -
Operated Fins, D. C. Friedman, Report OD-
1-171, NBS, Ordnance Development Division,

Feb. 26, 1944. Div. 4-222.123-MI

odd. Test to Compare Performance of Type S
(BRLG-6 Amplifier) and Standard MC-382
Units, R. Vorkink, Report OD-1-189, NBS,
Ordnance Development Division, Mar. 8,

1944. Div. 4-222.128-M31
See. Comparison of Performance of Type S and

Standard MC-382 on Motors with Scallop-
Type Traps, R. Vorkink, Report OD-1-197,

NBS, Ordnance Development Division, Mar.

15, 1944. Div. 4-222. 128-M32

off. Early Functioning of T5 Units, Tests of
Powder Lots, Motor Lots, Igniters, Traps,

D. W. Scott and T. N. White, Report OD-1-
227, revised Sept. 22, 1944.

Div. 4-222.121-M6
5 gg. Further Testing with Pellets and Salted
Powders, D. W. Scott, Report OD-1-241,

NBS, Ordnance Development Division, Apr.

18, 1944. Div. 4-222. 128-M33

Shh. Effect of Trap Structure on Early Function-
ing of T 5 Fuzes, D. W. Scott, Report OD-1-
253, NBS, Ordnance Development Division,

Apr. 22, 1944. Div. 4-222.128-M34

5ii. Effect of Fins on Mai-Functioning of T6
Fuze, D. W. Scott, Report OD-1-259, NBS,
Ordnance Development Division, Apr. 25,

1944. Div. 4-222. 122-MI

5jj. Effect of Salted Powder on Performance of
MC-382 Fuzes, D. W. Scott, Report OD-1-
274, NBS, Ordnance Development Division,

May 3, 1944. Div. 4-222.128-M35

5kk. Test of T6 Fuzes on Rigid-Fin Projectiles, 7.
D. W. Scott, Report OD-1-280, NBS, Ord-
nance Development Division, May 15, 1944.

Div. 4-222.122-M2 8.

511. Effect of Notched-Powder Loads on MC-382

Functioning , D. W. Scott, Report OD-1-287,
NBS, Ordnance Development Division, May
27, 1944. Div. 4-222.128-M36

5mm. Field Test of T5 on Projectiles with Bubble-
Wire Traps, D. W. Scott, Report OD-1-368,
NBS, Ordnance Development Division, June
19, 1944. Div. 4-222. 121-MI

5nn. Test of T 5 and T6 on Projectiles with Loose
Joints, D. W. Scott, Report OD-1-395, NBS,
Ordnance Development Division, July 8,
1944. Div. 4-222.123-M2

5oo. Test of T5 on Projectiles with Crimped and
Brazed Fins, D. W. Scott, Report OD-1-403,
NBS, Ordnance Development Division, July
17, 1944. Div. 4-222. 121-M3

5pp. Test of T 5 on Projectiles with Salted Powder
and Bubble-Wire Traps, D. W. Scott, Report
OD-1-397, NBS, Ordnance Development Di-
vision, July 10, 1944. Div. 4-222.121-M2
5qq. Performance of T 6 with 10 A Amplifiers on
M9, M9A1 and M9A2 Motors, D. W. Scott,
Report OD-1-404, NBS, Ordnance Develop-
ment Division, July 21, 1944.

Div. 4-222. 128-M37
5rr. Effect of Bayonet and Bag Igniters on Func-
tioning of T5 Fuze, D. W. Scott, Report OD-
1-408, NBS, Ordnance Development Di-
vision, July 19, 1944. Div. 4-222.121-M4
5ss. High-Angle Test of T 5 with 10 A Amplifier
(Some Shaker-Tested) on Motors with
Hand-Crimped Fins, Straightened and Un-
straightened, D. W. Scott, Report OD-1-423,
NBS, Ordnance Development Division, July
27, 1944. Div. 4-222.128-M38

5tt. Performance of Shaker-Tested T5 with 10 A
Amplifier, D. W. Scott, Report OD-1-477,
NBS, Ordnance Development Division, Aug.
24, 1944. Div. 4-222.128-M39

5uu. T5 on M9A1 with Clamp-On Fixed Fins,
D. W. Scott, Report OD-1-486, NBS, Ord-
nance Development Division, Sept. 5, 1944.

Div. 4-222. 121-M5
5vv. Flight Test of T 5 Fuzes on T22 Rockets with
EJA Propellant, D. W. Scott, Report OD-1-
614, NBS, Ordnance Development Division,
Jan. 4, 1945. Div. 4-222.121-M8

5xx. Effect of Trap Length on Incidence of Early
Functions in the T5, D. W. Scott, Report
OD-1-691, NBS, Ordnance Development Di-
vision, Mar. 29, 1945. Div. 4-222.121-M12
6. Salvo Firing in Search of Sympathetic Function-
ing of the MC-382, T. N. White, Report OD-1-15,
NBS, Ordnance Development Division, Sept. 25,
1943. Div. 4-245-MI

Rotation of M9A1 with Hand-Crimped Fins, D. W.
Scott, Report OD-1-588, NBS, Ordnance Develop-
ment Division, Dec. 18, 1944. Div. 4-412. 2-M3
Effect of Rocket Spin upon the Performance of
VT Fuzes TU, T5, T6, Theodore B. Godfrey, Re-

460

BIBLIOGRAPHY

port OD-1-668, NBS, Ordnance Development Di-
vision, Mar. 13, 1945. Div. 4-222.123-M3

9. Effect of Rain upon the Performance of VT
Fuzes , T5 and T6, Theodore B. Godfrey, Report
OD-1-669, NBS, Ordnance Development Division,
Mar. 13, 1945. Div. 4-222.123-M4

10. Effect of Rocket Spin on T5 Performance , D. W.
Scott, Report OD-1-677, NBS, Ordnance Develop-
ment Division, Mar. 21, 1945. Div. 4-222.121-M9

11. Effect of Rocket Spin on T5 Arming Distance,
D. W. Scott, Report OD-1-678, NBS, Ordnance
Development Division, Mar. 21, 1945.

Div. 4-222. 121-M10

12. Arming Time of T5 on T22 Fired Spiral Launcher,
D. W. Scott, Report OD-1-689, NBS, Ordnance De-
velopment Division, Mar. 28, 1945.

Div. 4-222. 121-Mil

13. Effect of Rotation Upon the Operation of the SW-

230 Switch, Charles C. Gordon, Report OD-1-729,
NBS, Ordnance Development Division, Apr. 30,
1945. Div. 4-238. 511-M6

14. Ballistic Test, MU3C Shell with Various Fuzes, G.
Rabinow, Report OD-1-737, NBS, Ordnance De-
velopment Division, May 8, 1945. Div. 4-514-M3

15. Field Test of 1U0 Philco T91 and 120 Emerson T92
— Various Release Conditions, Army Ordnance
Test, R. Vorkink, Report OD-1-825, NBS, Ord-
nance Development Division, July 9, 1945.

Div. 4-222. 114-M2

16. High-Angle and Target Test of T5 and T50 on T22
Rockets Modified for Helical Launcher, B. M.
Bennett, Report OD-1-895, NBS, Ordnance De-
velopment Division, Oct. 8, 1945.

Div. 4-222. 129-M5

17. Afterburning from Rocket Motors and Malfunc-
tioning of VT Fuzes (Summary Report), H. F.
Stimson, Report OD-1-896, Oct. 15, 1945.

Div. 4-411. 11-M6

18. Summary of Experimental Field Test Results of
Bomb Fuzes by Test Request Number, Ordnance
Analytical Group, Report OD-2-224, NBS, Ord-
nance Development Division. Div. 4-222.1 1-Ml

19. Summary of Pre-Production Mortar Fuze Field
Test Results, Ordnance Analytical Group, Report
OD-2-229, NBS, Ordnance Development Division,
June 18 and Sept. 27, 1945. Div. 4-222. 131-M4

20. Mortar Fuze Arming Time Tests, Ordnance Ana-
lytical Group, Report OD-2-230, NBS, Ordnance
Development Division, June 23 and July 14, 1945.

Div. 4-222.131-M5

21. Summary of Rocket Fuze Plane Firing (Air-to-

Earth) Tests, Ordnance Analytical Group, Re-
port OD-2-269, NBS, Ordnance Development Di-
vision, Aug. 24, 1945. Div. 4-222.126-M3

22. Arming of VT Fuzes: Analysis and Measurement

of Spread in Air-Trav el-to- Arming , A. L. Leiner,
Report OD-2-275, NBS, Ordnance Development Di-
vision, Mar. 14, 1946. Div. 4-244.1-M3

23. Arming Considerations in T6, Bertrand J. Miller

and Philip R. Karr, Report OD-3-74, NBS, Ord-
nance Development Division, Jan. 22, 1944.

Div. 4-238.515-MI

24. Minimum Useful Range for T6, Robert D.
Huntoon, Report OD-3-98, NBS, Ordnance De-
velopment Division, Feb. 9, 1944.

Div. 4-238. 515-M2

25. Computation of Burst Heights of Longitudinally -

Excited Bomb Fuzes. R. B. Schwartz, Report OD-
3-281, NBS, Ordnance Development Division, Aug.
7, 1945. Div. 4-241-M8

26. Summary of Rocket Fuze Experimental Field Test
Results, Analytical Group, Paul F. Bartunek and
C. F. Smolen, Report OD-7-97M, NBS, Ordnance
Development Division, Apr. 2, 1945.

Div. 4-222. 12-MI

27. Summary of Recent Target Tests at Blossom

Point, Alex Orden and C. F. Smolen, Report OD-
7-98, NBS, Ordnance Development Division, Apr.
9, 1945. Div. 4-222. 12-M2

28. Analysis of Variations in the Spread of Air-

Trav el-to- Arming , B. M. Bennett, Report OD-7-
103, NBS, Ordnance Development Division, Apr.
11, 1945. Div. 4-244.1-M2

29. Summary of Tests of T30 and T200U Rocket Fuzes

during the Period November 30, 19 UU to March
31, 19 U5, Paul F. Bartunek, Report OD-7-108,
NBS, Ordnance Development Division, Apr. 30,
1945. Div. 4-222.126-MI

30. Mortar Fuze Field Test Results, Experimental

Tests by Test Request, Analytical Group, Paul F.
Bartunek and C. F. Smolen, Report OD-7-112,
NBS, Ordnance Development Division, Apr. 23,
1945. Div. 4-222. 133-Ml

31. Tests BJM-5 and BJM-6, Charles Ravitsky,
Progress Report OD-7-206R, NBS, Ordnance De-
velopment Division, May 14, 1945.

Div. 4-222. 124-M3

MEMORANDA OF ORDNANCE DEVELOPMENT
DIVISION, NBS

32. Mid-Functioning, Section 1 Memo to Harry M.
Diamond from H. F. Stimson, NBS, Ordnance
Development Division, June 5, 1944.

Div. 4-222. 122-M3

33. Prediction of T51 Burst Height, D. 1 . Worcester
Technical Memorandum OD-l-TM-11, NBS, Ord-
nance Development Division, Nov. 8, l$4o

Div. 4-241-M9

34. Relation between the Spread in Burst Heights

and the Mean Burst Height of VT Bomb Fuzes,
R. C. Stillinger, Technical Memorandum OD-1-
TM-13, NBS, Ordnance Development Division,
Dec. 13, 1945. Div. 4-241-M10

35. Empirical Burst Height Distribution Formulae
for VT Bomb Fuze, R. C. Stillinger and Irene
Hess, Technical Memorandum OD-l-TM-23, NBS,
Ordnance Development Division, Sept. 17, 1946.

Div. 4-241-M11

BIBLIOGRAPHY

461

36. A Comparison of Observed and Predicted Burst

Heights of Ring-Type VT Bomb Fuzes, W. J.
Cronin and T. N. White, Technical Memorandum
OD-1-24M, NBS, Ordnance Development Division,
Sept. 19, 1946. Div. 4-241-M12

37. Analysis of T30 and T200U FOMA Tests, F. L.
Celauro, Memorandum OD-2-272M, NBS, Ord-
nance Development Division, Sept. 19, 1946.

Div. 4-126-M4

REPORTS OF CONTRACTORS OF DIVISION 3
(SECTION L) OF NDRC

38. Trajectories of Aircraft Rockets 3.5" and 5.0",

OSRD 2225, OEMsr-418, Service Projects OD-162,
OD-164, and NC-170, Division 3 Report CIT UBC
27, California Institute of Technology, Sept. 25,
1944. Div. 4-412. 1-M2

REPORTS OF CONTRACTORS OF
DIVISION 4, NDRC

39. Final Report: Summary Technical Paper, State

University of Iowa Staff, Contract OEMsr-769,
Sept. 29, 1945. Div. 4-100-M7

REPORTS OF APPLIED MATHEMATICS PANEL

40. Probability that a U.5" Rocket Fired from Astern

Will Destroy a Twin-Engine Bomber (Ju-88) as a
Function of the Point of Burst, Statistical Re-
search Group, Columbia University, AMP Report
21. 1R, July 1944, and Optimum Burst Surface for
U.5" Airborne Rocket Fired from Astern at Twin-
Engine Bomber (Ju-88), AMP Report 21.2R,
Statistical Research Group, Columbia University,
July 1944. Div. 4-412.3-MI, Div. 4-412.3-M2

41. Effectiveness of a U-5" Airborne Rocket with T5
Fuze when Fired at Twin-Engine Bomber from
Astern, AMP Report 21. 3R, Statistical Research
Group, Columbia University, July 1944.

Div. 4-412.3-M3

U. S. MILITARY PUBLICATIONS

Navy

42. Final Report on Air-to-Air Firing of Mk-171
Mod 0 Fuzes in 3.5" and 5.0" AR, NOTS Project
104 AFS, Serial 52, Aug. 5, 1945.

43. Final Report on Air-to-Ground Firing of Mk-172
Mod 0 Fuzes with 5.0" AR, NOTS Project 106 AF,
May 3, 1945.

Army

44. Proof Testing — A Brief Statistical Description of
Final Acceptance Sampling Formulas and Prov-
ing Ground Test Performance, W. Steele and E. J.
Fister, Camp Evans Signal Laboratory, Technical
Memo SA-Q1, Feb. 3, 1945.

45. Second Interim Report on Test of Fuzes, Bomb,
T50, Army Air Forces Board (Eglin Field,

Florida), Project F4012 (Test S.T. 1-44-12), Mar.
29, 1945.

46. Final Report on Test of Napalm-Gasoline Filled
M-10 Tanks with T50 and T51 Fuzes for Use
as an Incendiary Bomb, Army Air Forces Board
(Eglin Field, Florida), Project F4222 (Test S.T.

1- 44-91), Apr. 20, 1945.

47. Final Report on Comparison of the Effectiveness
of Bombs against Enemy Installations, Army Air
Forces Board (Eglin Field, Florida), Project
F4475 (Test S.T. 1-45-19), May 14, 1945.

48. Supplemental Test on Aircraft Rockets for Anti-
Personnel Effect, Army Air Forces Board (Eglin
Field, Florida), Project 4514 C471.94 (Test S.T.

2- 45-16) Sept 4, 1945.

49. Test of Fuze, Bomb, Nose T51E1, Army Air Forces
(Eglin Field, Florida), S.T.P 1-45-6, Nov. 27,
1945.

50. First Partial Report of Test of U-5-Inch Rockets
and Rocket Launcher, T-18E2 and T-20, Field
Artillery Board, Fort Bragg, N. C., Jan. 17, 1944.

51. Procedure for Conducting Field Engineering Ac-
ceptance Tests of Metal Parts Assentblies of VT
Bomb Fuzes, L. L. Parker, Apr. 3, 1945.

52. Tentative Specification, Ordnance Department,
AXS-1603, May 9, 1945.

53. Procedure for Conducting Field Engineering Ac-
ceptance Tests of Metal Parts Assemblies of VT
Fuze T200U, Army Ordnance Specifications, May
11, 1945.

54. Procedure for Conducting Field Engineering Ac-
ceptance Tests of Metal Parts Assemblies of VT
Fuze T200U, Army Ordnance Specifications, Aug.
3, 1945.

55. Army Ordnance Specifications AXS-1603 (Re-
vision 1), Aug. 13, 1945.

SUBDIVISION OF REFERENCES

56. In reference 26 the following test numbers: TBG-
93, -111, -115, -122A, -122B, -127, -130C, RQ-1C.

57. In reference 26 the following test numbers: TBG-
114, -120, -123, -124, -125, -128, RQ1A, RX2.

58. In reference 26 the following test numbers: TBG-
80, -90, -105, -131, -132, RX1A.

59. In reference 26 the following test numbers: TBG-
91, -94, -101, -105.

60. In reference 26 the following test numbers: TBG-
82, -86, -87, -88, -93, -103.

61. In reference 26 the following test numbers: TBG-
107, -108, -109A, -110, -113, -116, -117, -121, -126,
-130A.

62. In reference 26 the following test number: TBG-
130B.

63. In reference 26 the following test number: TBG-
109B.

64. In reference 26 the following test number: TBG-
113.

64a. In reference 26 the following test number:

TBG-85B.

462

BIBLIOGRAPHY

64b. In reference 26 the following test numbers:

TBG-112, 0118, -119.

65. In reference 18 the following test numbers: CB-

257, -270, -271, -272, -285, -301, -323, -324, -341,

-358, -389, -396, -414, -419, -420, -423, -431, -434,

-435, -445, -446, -448, -451, -452, -453, -454, -458,

-460, -462, -472, -474, -478, -483, -499, -500, -508,

-509, -518, -523, -527, SC-5, -9, -10, -12, -13, -14,
-15, -16, -17, -19, -20, PX-10, BX-4, -5.

66. In reference 18 the following test numbers: CB-
468, -491, -493, -501, -504, -506, -511, -515, -517,
-526, BX-1.

67. In reference 18 the following test number: CB-
506.

68. In reference 18 the following test numbers: CB-
468, -475, -491, -501, -515, BX-1.

69. In reference 18 the following test numbers: CB-
493, -504, -511, -517, BX-1.

70. In reference 18 the following test numbers: CB-
410, -416, -476, -502, PX-5, SC-7.

71. In reference 18 the following test numbers: CB-
344, -354, -376, -386, -489, -522, CEX-5, -7, BX-6.

72. In reference 18 the following test numbers: CB-
482, -485, -486, -487, -495.

73. In reference 18 the following test numbers: CB-
266, -349, -359, LSP-1, Ordnance Test (Aberdeen).

74. In reference 18 the following test numbers: CB-
487, -495, -514, -516, -522, BX-6.

75. In reference 18 the following test numbers: CB-
485, -486, -489, -495.

76. In reference 18 the following test numbers: CB-
475, -476, -489, -502.

77. In reference 18 the following test numbers: CB-
283, -284, -286, -289, -290, -291, -292, -294, -295,
-297, -299, -300, -304, -305, -307, -309, -310, -311,
-313, -314, -326, -327, -328, CHP-20, -21.

78. In reference 18 the following test numbers: CB-
475, -476, -482, -485, -486, -487, -495, -502, -515.

79. In reference 18 the following test numbers: CB-
344, -354, -386, -410, -416, -468, -482, -485, -486,
-487, -489, -491, -493, -495, -501, -504, -506, -511,
-515, -517, -522, -526, PX-5, SC-6, -7, CEX-5, -7,
BX-1, -6.

80. In reference 18 the following test numbers: CB-
266, -349, -359, Dahlgren test of T91, Ordnance
Test (Aberdeen).

81. In reference 18 the following test numbers: CB-

357, -360, -365, -369, -370, -371, -372, -373, -377,
-378, -379, -387, -388, -393, -402, -404, -411, -412,

-413, -417, -418, -421, -429, -430, -447, -450, -451,

-455, -458, -459, -461, -463, -465, -469, -473, -477,

-480, -488, -492, -497, -498, -500, -503, -513, -518,

-521, BX-1, -3, -7, -8, -11, -13, -15, -17, -18.

82. In reference 18 the following test numbers: CB-
-458, -459, -469, -477, -479, -480, -488, -492, -498,
-513, -518, -521, BX-3, -8, -11, -15, -18.

83. In reference 18 the following test numbers: CB-
369, -370, -371, -372, -425, -429, -447, -450, -451,
-457, -461, -463, -464, -465, -481, -500, -503, -510,
BX-1, -7, -16, -17.

84. In reference 18 the following test numbers: CB-
497, -500, -510.

85. In reference 18 the following test numbers: CB-

357, -360, -365, -370, -371, -372, -373, -381, -387,

-388, -393, -398, -399, -401, -404, -411, -413, -417,

-429, -447, -450, -451, -455, -458, -461, -465, -469,

-473, -477, -480, -481, -488, -492, -497, -498, -500,

-503, -513, -518, -529, BX-1, -3, -7, -8, -11, -15, -17,
-18.

86. In reference 18 the following test numbers: CB-
473, -480, -492, -496, -497, -498, -500, -507, -513,
-519, -525, BX-1, -3, -7, -8, -14, -15, -18.

MISCELLANEOUS REFERENCES PERTAINING
TO EFFECTIVENESS OF PROXIMITY FUZES

87. “Trials with AN-M.64 Bombs, Nose Initiated
(T-50) against Close Support Targets,” Ordnance
Board Proceedings No. Q2881, E. S. Pearson and
B. L. Welch, Dec. 13, 1944.

88. “Bombs, Aircraft, and Fuzes, Bomb, Aircraft: (1)
Report on test conducted in U. S. A.; (2) Theo-
retical calculations on optimum height of burst of
aircraft bombs fitted with V.T. fuzes.” Ordnance
Board Proceedings No. Q3860; notes, E. S. Pear-
son, Oct. 29, 1945.

89. Airburst for Blast Bombs, E. B. Wilson, Jr.,
NDRC Report A-322, April 1945.

90. Effect of Height of Detonation of Bombs on the
Blast Pressures and Impulses of Surrounding
Buildings, in Richmond Park 1/7 Square Model
Town Tests, Road Research Laboratory, Depart-
ment of Scientific and Industrial Research, Min-
istry of Supply, Note No. MOS/434/RJ.EK,
March 1945.

91. “Air Burst Bombs,” Memorandum from A. H.
Taub (Division 2, NDRC) to Col. P. Schwartz
(Director of Armament, USSTAF), Dec. 21, 1944.

92. “Air Burst Bombs — Status, as of 20 October
1944,” D. G. Christopherson, Ministry of Home
Security, REN-461.

93. Note on Airbursts of 4,000-lb. H.C. Bomb with
T-51 Fuze, F. H. East, Technical Note No. ARM-
343, Royal Aircraft Establishment, April 1946.

94. Interim Report, February 15 to March 7, 1945,
A. V. Astin to Dr. Alexander Ellett.

95. Inflammability of Mustard Chargings in British
Bombs A/C LC 500-lb Mark II Equipped with T-51
Fuzes, San Jose Project Report 71, June 23, 1945.

96. Optimum Height of Setting for T-50 Fuze on Blast
Bombs, A/C LC 500-lb Mark II Charged Dyed
Methyl Scelicyliate and Dropped onto Jungle,
San Jose, Project Report 69, Chemical Warfare
Service, June 22, 1945.

97. Multiple Bomb Assessment of Blast Bomb A/C
LC 500-lb Mark II Fitted with T-51 Fuze and
Charged HT When Dropped from High Altitudes
into Jungle Terrain, San Jose, Project Report 73,
Chemical Warfare Service, July 28, 1945.

98. Statistical Tables for Biological, Agricultural and
Medical Research, Fisher and Yates.

SECRET^

OSRD APPOINTEES

division 4

Chief

Alexander Ellett

Technical Aides

A. S. Clarke John S. Rinehart

Sebastian Karrer E. R. Shaeffer

Cathryn Pike A. G. Thomas

R. M. Zabel

Members

L. J. Briggs Harry Diamond

W. D. COOLIDGE F. L. HOVDE

J. T. Tate

Special Assistants

M. G. Domsitz
W. E. Elliott
Wendell Gould
W. S. Hinman, Jr.

Joseph Kaufman
J. L. Thomas

E. A. Turner

F. C. Wood

Consultants

A. V. Astin
R. A. Becker
R. M. Bowie
Cledo Brunetti
J. W. DuMond
Saul Dushman
Wm. Fondiller
T. B. Godfrey
L. R. Hafstad
J. E. Henderson
R. D. Huntoon
J. A. Jacobs
R. B. Janes
T. Lauritsen

D. H. Loughridge
W. B. McLean
F. L. Mohler
S. H. Neddermeyer
H. F. Olsen
C. H. Page
W. J. Shackelton

F. B. SlLSBEE

K. D. Smith

G. W. Stewart
J. F. Streib

L. S. Taylor
G. W. Vinal
W. L. Whitson

R. M. Zabel

CONTRACT NUMBERS, CONTRACTORS, AND SUBJECT OF CONTRACTS

Contract

Number

Name and Address of Contractor*

Subject

OEMsr-258

Friez Instrument Division, Bendix Aviation
Corporation

Baltimore, Maryland

Studies and experimental investigations in
connection with continuous development
work on special radio devices.

OEMsr-343

Westinghouse Electric and Manufacturing
Company

Baltimore, Maryland

Studies and experimental investigations in
connection with the development of
special radio devices.

OEMsr-500

Western Electric Company, Inc.

New York, New York

Studies and experimental investigations in
connection with the development of elec-
tronic devices.

OEMsr-528

National Carbon Company, Inc.

New York, New York

Production of small batteries suitable for
operation at low temperatures.

OEMsr-611

General Electric Company

Schenectady, New York

Studies and experimental investigations in
connection with the development of mini-
ature vacuum tubes, and report the re-
sults thereof.

OEMsr-566

Raytheon Production Corporation

Newton, Massachusetts

Studies and experimental investigations in
connection with the development of mini-
ature vacuum tubes.

OEMsr-630

Sylvania Electric Products, Inc.

Salem, Massachusetts

Studies and experimental investigations in
connection with the development of mini-
ature vacuum tubes having a very low
microphonic output.

OEMsr-769

University of Iowa

Iowa City, Iowa

Studies and experimental investigations in
connection with development work on
special electronic devices and associated
equipment.

OEMsr-866

Philco Corporation

Philadelphia, Pennsylvania

Studies and experimental investigations in
connection with the development of
special radio devices and associated equip-
ment.

OEMsr-885

Emerson Radio and Phonograph Corpora-
tion

New York, New York

Studies and experimental investigations in
connection with and carry on continuous
development work on special radio de-
vices and associated equipment.

OEMsr-887

Washington Institute of Technology
Washington, D. C.

Development of accessories for special elec-
tronic devices and associated equipment.

OEMsr-905

Western Electric Company, Inc.

New York, New York

Studies and experimental investigations in
connection with the development of
special electronic devices.

OEMsr-941

Federal Telephone and Radio Corporation
East Newark, New Jersey

Studies and experimental investigations in
connection with the development of
special selenium rectifiers.

* The National Bureau of Standards, which served as the central laboratories for Division 4, NDRC, did not operate under a contract
but as a government agency on a direct transfer of funds from OSRD.

CONTRACT NUMBERS, CONTRACTORS, AND SUBJECT OF CONTRACTS (Continued)

Contract

Number

Name and Address of Contractor *

Subject

OEMsr-949

University of Florida

Gainesville, Florida

Conduct theoretical studies and experi-
mental investigations in connection with
problems peculiar to special electronic de-
vices for ordnance application.

OEMsr-954

The Zell Corporation

Baltimore, Maryland

Furnishing machining facilities in connec-
tion with development of special elec-
tronic devices.

OEMsr-980

Zenith Radio Corporation

Chicago, Illinois

Studies and experimental investigations in
connection with development of special
electronic devices.

OEMsr-981

Knapp-Monarch Company

St. Louis, Missouri

Studies and experimental investigations in
connection with the development of
special power supplies and associated
equipment.

OEMsr-1003

Radio Corporation of America

Harrison, New Jersey

Studies and experimental investigations in
connection with development of special
miniature vacuum tubes.

OEMsr-1106

Westinghouse Electric and Manufacturing-
Company

Washington, D. C.

Pilot production of special electronic de-
vices.

OEMsr-1109

General Electric Company

Schenectady, New York

Studies and experimental investigations in
connection with development work on
special electrical and radio devices and
associated equipment.

OEMsr-1113

Emerson Radio and Phonograph Corpora-
tion

New York, New York

Manufacture and delivery of special elec-
tronic devices.

OEMsr-1117

Globe-Union, Inc.

Milwaukee, Wisconsin

Studies and experimental investigations in
connection with development of special
electrical and mechanical devices.

OEMsr-1133

Zenith Radio Corporation

Chicago, Illinois

Manufacture and delivery of special elec-
tronic devices.

OEMsr-1134

Knapp-Monarch Company

St. Louis, Missouri

Manufacture and delivery of special power
supplies.

OEMsr-1161

The Rudolph Wurlitzer Company

North Tonawanda, New York

Studies and experimental investigations in
connection with the development of
special electronic devices.

OEMsr-1163

The Rudolph Wurlitzer Company

North Tonawanda, New York

Manufacture and delivery of special elec-
tronic devices.

OEMsr-1196

Philco Corporation

Philadelphia, Pennsylvania

Manufacture and delivery of special elec-
tronic devices.

SECRET

465

CONTRACT NUMBERS, CONTRACTORS, AND SUBJECT OF CONTRACTS (Continued)

Contract

Number

Name and Address of Contractor*

Subject

OEMsr-1227

Bowen and Company, Inc.

Bethesda, Maryland

Furnish necessary machine shop and assem-
bly facilities for the development of
special electronic devices.

OEMsr-1251

General Electric Company

Schenectady, New York

Manufacture and delivery of special elec-
tronic devices.

OEMsr-1378

Raymond Engineering Laboratory

Berlin, Connecticut

Studies and experimental investigations in
connection with development of special
electronic devices.

OEMsr-1437

The General Instrument Corporation

Elizabeth, New Jersey

Studies and experimental investigations in
connection with development of electrical
and mechanical devices.

OEMsr-1477

Zenith Radio Corporation

Chicago, Illinois

Development and production of special
electronic devices.

OEMsr-1500

Emerson Radio and Phonograph Corpora-
tion

New York, New York

OEMsr-1501

Solar Aircraft Company

San Diego, California

Design and produce donut-type setback
arming devices for use on British rockets
equipped with VT fuzes.

466

SECRET

SERVICE PROJECT NUMBERS

The projects listed below were transmitted to the Executive
Secretary, NDRC, from the War or Navy Department through
either the War Department Liaison Officer for NDRC or the
Office of Research and Inventions (formerly the Coordinator of
Research and Development), Navy Department.

Service

Project

Number

Subject

Chemical Warfare Service

CWS-19 Development of an influence fuze for airplane spray apparatus.
Ordnance Department

OD-27 Development of proximity (influence) fuzes for bombs and
projectiles.

OD-191 Development of VT fuze and UHF and VHF circuit techniques.
OD-192 Development of counter-countermeasures for VT fuzes.

SC-38

SC-40

Signal Corps

Field testing equipment for proximity fuzes.
Substitute for dry battery BA-55.

467

INDEX

The subject indexes of all STR volumes are combined in a master index printed in a separate volume.

For access to the index volume consult the Army or Navy Agency listed on the reverse of the half-title page.

A-l mortar shell retrieving device, 355-
357

A-2 mortar shell retrieving device, 356-
357

Acceleration integrators for arming prox-
imity fuzes, 171-174
British type, 172
double action arming device, 173
for T-4, T-5 and T-6 fuzes, 172
Acceptance testing, requirements, 428-
432

Navy rocket fuzes, 430-431
T-5 fuzes, 431-432
VT bomb fuzes, 428-430
Active-type fuzes, 4-5
Afterburning in rockets
Ballistite burning, 364-365
burning process, 364-365
definition, 211

pellets added to Ballistite, 365
salted powder additions, 366
summary, 336

trap ring variations, 365-366
with T-30 fuze, 380-382
with various propellants, 337
Air-burst bombs, effectiveness, 412-416
against moderately shielded person-
nel, 412-413

against shielded personnel and un-
shielded materiel, 413-414
against unshielded materiel and en-
trenched personnel, 414
blast effect, 415-416
compared with contact-burst bombs,
414

spread of gas, 416
Air-burst fuzes, effectiveness, 14-15
Aircraft, reflecting properties, 61-64
effect of wavelength, 64
experimental measurement, 61-63
sensitivity requirements for plane-to-
plane rocket fuze, 63-64
Air-travel devices for safety in arming
fuzes, 169

Allis-Chalmers Company, fuze bear-
ings, 190

Alnico rotors for fuze generators, 269
Alternator, permanent magnet, 146-147
American Phenolic Corporation, ther-
mosetting cement, 248
Amphenol 912 cement, 248, 252
Amplifier systems, 103-117, 256-265
adjustment and testing, 115
ceramic amplifiers, 259
disk construction, 259
potting and impregnating, 261-265

properties of pentodes, 114-115
requirements, 103-104, 256
response to spurious signals, 115-116
ring construction, 259
sandwich construction, 257-258
tolerance of components and varia-
tion in performance, 116-117
Amplifier systems, characteristics, 104-
110

for airborne target, longitudinal ex-
citation, 105-106

for ground approach, longitudinal
excitation, 106-108
for ground approach, transverse ex-
citation, 108-110
Amplifier systems, gain, 110-114
axial antenna fuzes, 110-111
combination amplifiers, 113-114
gain-control condensers, 261
transverse-antenna fuzes, 112-113
Angle of approach, rocket, 339-340
Antenna

constant, evaluation, 75-77
fuze, axial, 110-111
fuze, transverse, 30-32
noise from propellant flames, 72-75
noise resulting from geometric de-
formations, 71-72

reflectors, use as signal simulators,
66-67

size limitations, 167
Antenna and target, interaction phe-
nomena, 22-24

fundamental equations, 22-23
fuze equations, 23-24
fuze problem as interaction of two-
terminal networks, 22
Antenna impedance, 17-22, 37-43
approximations in impedance repre-
sentation, 19-21

ground-approach case, magnitude and
frequency of impedance signal,
51-54

ground-approach case, prediction of
height of function, 50-51
impedance concept, 21-22
input impedance, 23
radiation resistance, 38-43
reactance across antenna terminals,
39-40, 41-43

reflected wave or doppler frequency
concept, 18

reflection equivalent to change of
antenna impedance, 18-19
specification of antenna terminals,
37-38

SECRET

Antenna impedance, airborne target,
59-64

properties of impedance signal wave,
59-61

reflecting properties of aircraft, 61-64
Antenna impedance modulation, cir-
cuit response, 34-37
differential signals, 34-35
finite signals, 35
fuze circuit parameters, 35-37
Antennas, transverse, 48-49
Antiaircraft use of fuzes, summary, 14-
15

Antimateriel bomb fuzes, military re-
quirements, 3
Apex firing test, 296
AR 5.0 Navy rocket fuze, 217-220
amplifier, 220
amplifier gain, 237
AR rocket, characteristics, 326
arming mechanism, 219
burst height, 219, 237
characteristics, 235
firing circuit, 220
limitations, 218
military requirements, 217
power supply, 220
radiation pattern, 237
release altitude, 236
r-f system, 220
safety and arming, 218
Arming methods

acceleration integrators, 171-174
air-travel devices, 169, 388, 393-394
arming delay, 212
arming pulse, 99, 296
arming wire, 169
clocks and timing devices, 170
dashpot arming device, 192
“doughnut” mechanism, 187
effect of air pressure, 170
effect of propellant temperature, 336
electric arming, 125-130
for accelerated projectiles, 212
for battery-powered rocket fuze, 159-
160

for bomb fuzes, 224-225, 321-322,
387-394

for mortar shell fuzes, 418-419
for nonaccelerated projectiles, 212
for rocket fuzes, 159-160, 333-335
manual arming, 169
RC arming, 125-129, 335
safety features, 212
Army 4.5-in. rocket fuzes
see 4.5-in. Army rocket fuzes

469

470

INDEX

Audio portion of fuze
function, 284
input circuits, 284-285
output circuits, 285-286
production testing, 302-304
signal simulator, 69-70
tests, 284-287
thyratron tests, 286-287

BA-55 battery pack, 136-138

Ball bearings for proximity fuzes, 178

Ballistics of rocketry

angle of approach, 339-340
rate of spin, 340
velocity and acceleration, 339
yaw, 340

Ballistite burning in rockets, 364-365,
380-381

Bar-type bomb fuzes, 221-228, 405-412
see also T-51 fuze; T-82 fuze
amplifier, 226

burst height, 224, 407, 411-412
delayed arming device, 410-411
description, 221-222
effect of release conditions, 408-409,
411

effect of vehicle, 408, 411, 412
guide plates, 410
properties, 228
reliability, 223
summary, 221, 407-408
testing conditions and devices, 282-
283, 406-407

train release, 408-409, 410
washers, 409-410
yellow carrier, 234
Batteries for fuzes
dry, 133, 136-138
reserve, 133-134, 138-140
vibrator, 134
Battery fuzes
see also T-5 fuze
arming mechanism, 176
detonators, 177
head, 176
MC-382; 92

mechanical design, 175-177
rocket fuzes, 158-160
switch contacts, 177
BC (battery command) telescope, 345
Bell Telephone Laboratories, P-4 771B
bomb fuze, 199
Bomb fuzes
amplifier, 226

arrangement of components, 224
burst heights, 224, 230
directional sensitivity, 6-7
firing circuit, 227

military requirements, 2-3, 220-221
oscillator assemblies, 225
P-4 771B fuze, 198

✓

plastic and metal vanes, 142-143
power supply, 227
production data, 228, 229
reliability, 223
r-f system, 222, 225
safety, 222-223, 225
specific applications, 222
train-bombing, 322-323
Bomb fuzes, arming, 387-394
air travel, 169, 388, 393-394
arming wire, 169
effect of bomb, 388
effect of plane speed, 388
effect of release altitude, 388
mean air travel vs\ rotor setting, 391-
393

MinSAT settings, 389-390
reasons for study, 388
release methods, 389
rotor setting, 388-389, 393-394
tests, 321-322, 390-391
vane speed variations, 393
Bomb fuzes, specific models

see also T-50 fuze; T-50E1 fuze;
T-50E4 fuze; T-51 fuze; T-82
fuze

T-40; 3, 198
T-43; 3, 198

T-51E1; 234, 409-410, 422-426
T-89; 10, 13, 229, 396-397
T-90; 13, 232, 398
T-91; 10, 13, 229, 397
T-91E1; 397-398
T-92; 10, 13, 232
T-92E1; 232, 398
Bomb fuzes, tests, 313-324
arming tests, 321-322
assembly of components, 314-315
bomb preparation, 314
bomb types tested, 313-314
dive tests, 323
function heights, 317-318
function time, 318-319
fuze carrier characteristics, 319-321
phonographic determination of func-
tion time, 319

photographic determination of func-
tion heights, 317-318
plane-to-ground communications, 315
purpose, 313
range layout, 315-316
time lags, determination, 323-324
train tests, 322-323
Bomb fuzes, types
air-burst bomb fuzes, 412-416
bar-type, 221-228, 405-412
general types, 221
generator-powered fuzes, 160-165
ring-type, 223-229, 232, 394-405
vane types, 223

British

acceleration-operated arming device
for fuzes, 172

HC (high-capacity) bombs, 415-416
Brown carrier fuzes
amplifier gain, 229
bomb fuzes, 229
burst heights, 230
performance, 400
radiation patterns, 231
ring-type, 235
rocket fuzes, 240
“Burst,” definition, 361

Cementing of tubes in fuzes, 252
Cements

for anchoring fuze parts, 206
for fuze oscillators, 248
Cenco rocket, 326-327
Centrifuge for large fuzes, 298
Ceramic amplifiers, 259
Ceramic oscillator .blocks, 250, 253-256
assembly, 256
construction, 253
electrical properties, 253
mechanical properties, 253
metalizing, 254
resistoring, 255

soldering to ceramic surfaces, 255
Chemical-bomb fuzes, 3
Chemical- warfare spray tank, 10
CIMA fuzes, performance, 418-419
Clock mechanism for arming fuzes, 170,
192

Construction of proximity fuzes
see Mechanical design of fuzes
Contact-burst bombs, compared with
air-burst bombs, 414
Copper oxide rectifiers, 154-155
Critical voltage, definition, 286

Dashpot arming device for fuzes, 192
Definitions, 211
Detector circuit, 43-44
Detonation of fuzes, 213
impact detonation, 175
in-line detonators, 169
Detonator circuit, 117-131
capacitor, 120-122
detonator, 118-120
electric (RC) arming, 125-130
firing system, 155-156
operation, 124-125
requirements, 117-118
safety features, 130-131
self-destruction, 131
tetryl booster, 118
thyratron, 122-124
time lags, 119-120

Developmental relations among fuzes,
210-211

INDEX

471

Diode tube tests, 291
Dipole

see Antenna

Directivity patterns, 43-50
errors due to ground reflection, 77-S0
measurement, 43-45
reflections from ground, 44-45
space radiation pattern, 25
Directivity patterns, longitudinal ex-
citation, 45-48

comparison of patterns, 47-48
general features, 46-47
typical patterns, 45-46
Directivity patterns, transverse excita-
tion, 48-50

loop excitation, 49-50
transverse dipole, 48-49
Dive bombing tests, VT fuzed bombs,
323

“Dog collar’’ construction of amplifier,
259

Doppler frequency

antenna impedance, 18
reflected impedance, 33
Doppler fuzes, 4-9

operation and principal components,
5-9

optimum burst height, 7-8
Douglas Aircraft Company, nozzles for
fuzes, 200

Dow potting materials, 208
Dow Q247 plastic for fuzes, 205, 206
Dynamic balancing of proximity fuzes,
168, 201-203

Dynamic torque tests, 299

Early functioning
see Afterburning

Effective critical voltage, definition, 286
Electronic systems, 81-166
amplifier, 103-117
detonator circuit, 117-131
power supplies, 131-157
radio-frequency unit, 81-103
Electronic systems, coordination, 157-
166

battery-powered rocket fuze, 158-160
generator-powered bomb fuze, bar
type, 164-165

generator-powered bomb fuze, ring
type, 160-164

generator - powered trench - mortar
shell fuzes, 165-166

Field testing, 312-359
bomb fuzes, 313-324
mortar shell fuzes, 340-359
procedure and equipment, 312-313
purpose, 312
rocket fuzes, 324-340
Filter condensers for proximity fuzes, 27 0

5.0 Navy rocket fuze
see AR 5.0 Navy rocket fuze
FOMA fuzes, performance, 418-419
4.5-in. Army rocket fuzes, 213-217,
363-376

see also T-5 fuze; T-6 fuze
afterburning, 364-366
amplifier, 217
arming mechanism, 216
arrangement of components, 216
ground-to-ground firing, 216
limitation, 213
middle functioning, 366-367
military requirements, 213
plane-to-ground firing, 216
plane-to-plane firing, 215
radiation pattern, 238
rain effect, 368
r-f system, 217
safety and arming, 214
scoring methods, 363-364
self-destruction, 215
spin effect on arming, 368
sympathetic functioning, 367-368
Fragmentation bomb fuzes, military
requirements, 3

F ragmentation effect of air-burst bombs,
412-414

Fuze nomenclature, summary, 362
Fuze operation in flight, 319-321
generator speed, 320-321
mechanical trouble, 319-320
observational procedure, 319

Gain-control condensers for amplifiers,
261

Gas bombs, effectiveness, 416
Gauging tests, 300, 310
Generator

production testing, 303-305
speed, bomb fuze in flight, 320-321
speed, mortar shell fuzes, 351
storage systems, 134-135
testing, 294-295

Generator, construction, 267-270
bearings, 268
coil construction, 268
housing, 267
rotors, 269
shafts, 269

stator impregnation, 269
Generator, mechanically-driven rotary,
134

Generator, wind-driven, 135-136, 140-
153

alternator, 146-147
bearings, 144-145
dynamic balancing, 145
electric design, 145-150
nose-mounted vane, 140
operating range, 141

production models, 150-153
rotor, 148-150
single serpentine coil, 151
six-coil generator, 150-151
vane and turbine, 141-144
voltage regulation, 147-148
Generator-powered fuzes, 160-165, 177-
186

amplifier requirements, 162-163
antenna design, 162
arming, 164

bomb fuze, bar-type, 164-165
bomb fuze, ring-type, 160-164
carrier frequency, 161-162
feedback amplifier circuit, 111
miniature fuzes for trench mortars
and rockets, 188-198
oscillator-diode circuit, 162
overall stability, 164
power supply, 163

RRLG fuze for rocket application,
177-178

size and location, 161
specific models, 179-188
trench-mortar, 165-166
Glidden PT1 and PT2 used for potting
fuzes, 208

Glider bomb fuzes, military require-
ments, 2-3

Globe-Union Company

arming mechanisms for battery fuzes,
177

ceramic oscillator blocks, 253-256
T-132 fuze, 189-193, 241-244, 417-
419

GP (general-purpose) bomb fuze, mili-
tary requirements, 2-3
Ground-approach fuzes

amplifier characteristics, 106-110
antenna impedance, 23-24, 50-54
reflected impedance, 28
summary of characteristics, 10-13

HC (high-capacity) bombs, British,
415-416

Humidity tests, 297-298
HVAR rocket, characteristics, 326

IE-28 test set, 298

Impact detonation of proximity fuzes,
175

Impedance, mutual, radiation field, 24-
27

antenna gain, 26

between two arbitrary antennas, 26-
27

field equations for arbitrary antenna,
25-26

Impedance, reflected, 27-34
airborne target equation, 28-30
general properties, 33-34

472

INDEX

ground interference, 30
ground-approach equation, 28
transverse antenna fuze, 30-32
Impedance antenna
see Antenna impedance
Impedance signal, 51-54, 59-61
see also Signal simulators
ballistic and target factors, 53-54
fuze antenna factors, 51-52
wave amplitude, 60-61
wave phase, 59-60

Inertia arming of proximity fuzes, 171-
174

In-line detonators for proximity fuzes,
169

Jamming fuzes, antenna impedance, 24
Jolt test for fuzes, 191, 297

Katrinka bomb fuzes, 198

Laboratory tests, 278-311
audio portions, 284-287
complete units, 295-296
component testing, 290-295
gauging, 300

mechanical tests, 298-300
overall test, 278

pilot production test line, 300-308
procedure, 278
purpose of tests, 278
quality control testing, 308-311
radio-frequency sections, 278-284
service tests, 297-298
stability, 287-290
Launchers for rocket testing, 327
LC (light-case) bomb fuze, military
requirements, 2-3

Loading devices for fuze testing, 280-283
Loading requirements of fuzes, 279-280
Longitudinally excited proximity fuze,
167

Lucite fuze caps, rain protection device,
368

M-2 electric detonator, 119
M-8 rocket fuze
see T-5 fuze

M-10 chemical warfare spray tank, 415
M-64 air-burst bomb, effectiveness,
412-416

M-81 air-burst bomb, effectiveness,
412-416
M-166 fuze
see T-51 fuze
M-168 bomb fuze, 229
Manual arming of proximity fuzes, 169
MC-382 rocket fuze
early functioning, 371
radiating system, 89
tube characteristics, 92

Mechanical design of fuzes, 167-208
arrangement of main components,
167

battery fuzes, 175-177
choice of plastics, 204-208
dynamic balancing, 201-203
experimental fuzes, 198-199
general requirements, 167-168
generator fuzes for rockets and bombs,
177-188

miniature fuzes for trench mortars
and rockets, 188-198
mounting of fuzes into missiles, 199-
200

rigidity, 168

safety and arming, 168

size, 168

speed regulation for windmills and
turbines, 200-201

“Michigan sensitivity” of a fuze, 64, 83
“Micro-Dynetric” balancing of fuze,
203

Military requirements, 1-4
arming and safety requirements, 2
functioning point, 1
mechanical features, 1-2
MinSAT (minimum safe air-travel-to-
arming), 222-223, 389-390
Mk-171 fuze
see T-30 fuze
Mk-172 fuze
see T-2004 fuze

Monsanto Styramic 18, plastic mate-
rial for fuzes, 205

Mortar shell fuzes, 188-197, 340-359
see also T-132 fuze; T-171 fuze; T-172
fuze

arming, 418-419

breech-loading mortar recovery, 359
dynamic balancing, 202
electronic system, 165-166
firing coordination, 342-343
fuze flight performance, 350-351
gun position, 344-346
height of function, 350, 353-356
loading operations, 343-344
mortar shell trajectories, 352-353
packaging tests, 417-418
performance under standard condi-
tions, 416-417
ranges, 419-420
safety in arming, 171
test measurements, 346-350
weather effects, 351-352
Mortar shell retrieving devices, 355-357
Mortar shell trajectory calculation,
352-353

Mounting of fuzes into missiles, 199-
200

Mustard gas air-burst bombs, 416

Napalm-gasoline gel, fire bombing use,
414-415

National Bureau of Standards

acceleration integrator for arming
fuzes, 171-174

bearings for T-132 and T-171 fuzes,
190

centrifuges for fuze tests, 177, 188
dipoles for fuzes, 184
fuze battery, 133-134
gear train, 186
T-12 fuze, 177-178
T-171 fuze, 188-195, 241-244, 418-419
testing equipment for proximity
fuzes, 275

Navy rocket fuzes, 377-386
see also AR 5.0 Navy rocket fuze;

T-30 fuze; T-2004 fuze
acceptance testing requirements, 430-
431

arming distances, 378-379
dumping, 379
general discussion, 377
mechanical arming, 377-378
pulsing tests, 378-379
safety tests, 380
Noise antenna, 71-75
Noise sources in fuzes, 287-290
Nomograph for use in mortar shell test-
ing, 348-349

Normal critical voltage, definition, 286
Nose assembly of proximity fuzes, 265-
266

Nose fuzes

see T-50 fuze; T-51 fuze
NR-3A Raytheon tube, characteristics,
92-93

NS-3 Sylvania tube, characteristics, 92

OD (oscillator-diode) fuzes, 100
Oscillators, 247-256
carrier frequency uniformity, 250
ceramic blocks, 250, 253-256
coil construction, 250-252
design, 88-89
metalizing of blocks, 254
power oscillating detector, 84-85
“printed” circuits, 250
production plant testing, 302
production procedures, 247-253
reaction grid detector, 84, 98
requirements, 247
thermoplastic blocks, 248
thermosetting phenolic blocks, 248
tube mounting, 252
types of construction, 248

P-4 bomb fuze, generator design, 141,
153

P-4 771B bomb fuze, 199
Packaging tests, 297

INDEX

473

Parachute recovery devices for VT
mortar fuzes, 355-359
Passive fuzes, 5

PD M-4 fuzes, comparison with T-6
fuze, 376

Pellets for elimination of afterburning
in rocket propellants, 365-366
Perchloric acid battery cell, 139
Performance terminology for fuzes, 211
Phenolic thermosetting oscillator
blocks, 248

Photographic observations
detonation of VT mortar shells, 353-
356

in bomb fuze testing, 317-318
Pilot plant production of fuzes, 245
Pilot production test line, 300-308
audio prepot and postpot test posi-
tions, 303-304

audio pretest position, 302-303
generator test position, 303-305
head test position, 303
oscillator pretest position, 302
performance testing, final, 306-307
power supply test position, 306
pulse test, 307-308
rectifier-filter test position, 305-306
Piton-Bressant method, mortar shell
trajectory calculations, 352-353
Plastics for fuzes, 204-208
basic requirements, 204
cementing, 206
solder flux, 208
thermoplastic materials, 204
POD (power oscillating detector), 82,
84-85

Pole-test measurement of fuze sensi-
tivity, 65, 87

Potting of amplifiers, 261-265

as part of the production line, 264
Glidden compound, 265
immersion in hot waxes, 262
in the fuze cavities, 263
ingredients, 263
tung oil mixtures, 264
vacuum potting, 264
Potting of fuzes, 207-208, 253
Powder train interrupters, 168
Power oscillating detector, 82, 84-85
Power radiation pattern of fuze antenna
see Directivity patterns
Power supplies, 131-157, 266-272
dry batteries, 133, 136-138
electric components, 270-272
filter and detonator firing system,
155-156

filter condenser, 270
generator, 134-136, 267-270
production testing, 306
rectifiers, 153-155, 271
requirements, 131-133, 267

reserve batteries, 133-134, 138-140
supply circuits, 156-157
survey of possible sources, 133-136
testing, 294-295, 306
types, 266

wind-driven generators, 140-153
Production of fuzes, 227-239, 245-277
achievement, 277
amplifiers, 256-265
assembly line, 275
bar-type fuzes, 228
nose assembly, 265-266
organization and planning, 245-247
oscillators, 247-256
pilot plant, 245

power supply and arming, 266-272
process flow chart, 246
production techniques, 272-275
ring-type fuzes, 228
soldering, 272
testing, 275-277, 300-308
VT bomb fuzes, 228
Proper function of a fuze, definition, 211
Proximity fuzes
effectiveness, 11-16
electronic systems, 81-166
field testing, 312-359
laboratory tests, 278-311
mechanical design, 167-208
military requirements and objectives,
1-4

performance, 360-432
production, 227-239, 245-277
radiation interaction system, 17-80
Proximity fuzes, types

see also Bomb fuzes; Rocket fuzes
active fuzes, 4-5
bar-type fuze, 221-228, 405-412
battery fuzes, 175-177
doppler fuzes, 4-9

generator-powered fuzes, 160-165,
177-186

mortar shell fuzes, 188-197, 241-244,
340-359, 416-420

ring-type fuzes, 223-229, 394-405
Pulse test, 296, 307-308
“Purge pellets” for use in rocket propel-
lants, 365-366

Quality control testing, 308-311

comparison with pilot production
testing, 308-309
procedure, 309
specific tests, 310-311

Radiating system, 89-90
Radiation resistance

effect of feed geometry, 39-40
effect on reflected impedance, 33
experimental measurement, 38-39
typical values, 41-43

Radiation theory, 17-80
antenna impedance, 17-22, 37-43
antenna noise, 71-75
circuit response to antenna imped-
ance modulation, 34-37
directivity patterns, 43-50, 77-80
evaluation of antenna constant, 75-
77

mutual impedance, 24-27
reflected impedance, 27-34
signal simulation, 64-71
two-terminal networks, 22-24
working signals, airborne target, 59-
64

working signals, ground-approach
case, 50-54

Radiation theory, induction field, 54-59
effect on function heights, 57-59
second approximation to the field
equations, 55-57
Radio proximity fuzes
see Proximity fuzes

Radio rocket longitudinal generator
(RRLG), 177-178
Radio-frequency system, 81-103
carrier frequency, 284
loading devices, 280-283
loading requirements, 279-280
oscillator design, 88-89
power oscillating detector, 82
radiating system, 89-90
reaction grid detector, 82
requirements, 81-82
sensitivity, 82-89, 102-103, 283
shielding of fuzes, 280
spurious signals and circuit stability,
95-100

stability, 283
tests, 278-284
tube characteristics, 90-95
typical designs, 100-102
Radio-frequency system, signal simu-
lators
field, 65

laboratory, 65-69

resistance component simulators, 66-
68

rotating vector simulators, 68-69
Radius of action (ROA) of fuzes, defini-
tion, 211

Rain

effect on rocket fuzes, 337, 368
protection with Lucite fuze caps, 368
Random function of a fuze, definition,
211

Rate of spin, rocket, 340
Raymond Engineering Laboratories,
clock rotor for fuzes, 192
RC arming, 125-129
dumping, 128-129
measurement, 335

474

INDEX

pulse protection, 129
testing, 296

Reactance across antenna terminals
effect of feed geometry, 39-40
measurement, 39
typical values, 41-43
Reaction grid detector
circuit characteristics, 93
design, 88-89
dynamic stability, 98
performance compared with idealiza-
tion, 84

suggested antimicrophony circuit, 99
tuning effects, 100-101
Recordak viewers, use in bomb fuze
testing, 317-318, 320
Rectifier system

blocking layer rectifiers, 154-155
filters, production testing, 305-306
for proximity fuzes, 271-272
testing, 293-294
vacuum-tube rectifiers, 154
Reflection

see Radiation theory
Resistance component signal simu-
lators, 66-68
diode, 67-68
dipole reflectors, 66-67
dummy antenna, 66
thermistors, 68
triode, 68

Resistors, compensated, 281-283
R-f system

see Radio-frequency system
RGD oscillator

see Reaction grid detector
Rigidity of proximity fuzes, 168
Ring construction of amplifiers, 259
Ring-type bomb fuzes, 223-229, 394-405
see also T-60 fuze
amplifier, 226

arming devices, delayed, 402-403
brown carrier, 228, 229
burst heights, 224, 396, 398-399, 403-
405

comparison with bar-type fuzes, 221
effect of release altitude, 401, 403-404
effect of train release, 404-405
effect of train spacing, 401
effect of vehicle size, 400, 404
fin insulators, 402, 404
fin thickness, 402, 403-404
fuze protective devices, 401-402
production data, 229, 232
reliability, 223
washers, 401
white carrier, 228, 232
Ring-type bomb fuzes, acceptance tests,
394-405

burst height distribution character-
istics, 398-399

conditions for acceptance, 394
effect of test conditions on perform-
ance, 395-396
mean burst heights, 396
metal parts, 395
summary, 399-400
T-50-E1 fuze, 396-397
T-50-E4 fuze, 398
T-89 fuze, 396-397
T-90 fuze, 398
T-91 fuze, 397
T-91-E1 fuze, 397-398
T-92-E1 fuze, 398
Ring-type rocket fuzes
see AR 5.0 Navy locket fuze; T-30
fuze

ROA (radius of action) of fuzes, 211
Rocket ballistics
angle of approach, 339-340
rate of spin, 340
velocity and acceleration, 339
yaw, 340

Rocket fuzes, 235-241, 324-340
see also AR 5.0 Navy rocket fuze;

4.5-in. Army rocket fuzes
arming, 159-160, 333-335
ballistics of rockets, 339-340
carrier performance, 331-332
effect of propellant temperature, 336
effect of raindrops, 337-339
fin structure, 336
firing from airplane, 331
high-angle firing, 331
horizontal firing, 329-331
metal vanes, 143
ring-type, 235

rocket characteristics, 326-328
safety in arming, 171
sensitivity and burst surface, 332-333
sensitivity requirements, 6-7, 63-64
sympathetic functioning, 339
testing procedure and equipment,
324-326

water-approach tests, 332-333
Rocket fuzes, afterburning in
Ballistite burning, 364-365
burning process, 364-365
definition, 211

pellets added to Ballistite, 365
summary, 336

trap ring variations, 365-366
with various propellants, 337
Rocket fuzes, battery-powered
amplifier requirements, 159
arming, 159-160
carrier frequency, 158-159
mechanical stability, 160
oscillator and detector, 159
power supply, 159 ^

size and location, 158 *

Rocket fuzes, specific models

see also T-5 fuze; T-6 fuze; T-12 fuze;
T-30 fuze; T-2004 fuze; T-2005
fuze

Rotating vector signal simulators, 68-
69

Rotors for fuze generators, 269
RRLG fuze, 177-178

Safety requirements, 168-175
see also Arming methods
comparison of proximity fuzes with
other fuzes, 168
for detonator circuit, 130-131
for 4.5-in. Army rocket fuzes, 214
impact detonation, 175
powder train interrupters, 168
rotating and nonrotating projectiles,
169

self-destruction, 174
Salt spray tests, 298
Sandwich construction of amplifiers,
257

Selenium rectifiers, 154-155, 271-272,
274

Self destruction (SD) mechanism of
fuzes, 131, 174, 215, 296
Sensitivity of fuze, 82-89
definition, 82-87
directional sensitivity, 6-7
experimental determination, 87-88
“Michigan sensitivity,” 64, 83
pole-test measurement, 65, 87
radio-frequency sections, 283
rocket fuze sensitivity, 63-64, 332-
333

sensitivity concept, 102-103
Signal simulators, 64-71
field r-f simulator, 65
laboratory audio simulator, 69-70
laboratory r-f simulators, 65-69
overall signal simulator, 71
required properties, 64-65
Signals, fuze
differential, 34-35
finite, 35

Signals, spurious, 95-100

antimicrophony circuits, 98-99
arming pulse, 99
component noise, 95-96
corona effects, 96
response of amplifier, 115-116
unstable oscillation, 96-98
Size of proximity fuzes, 168
Solar Aircraft Company, doughnut
arming mechanism, 188
Soldering

ceramic surfaces, 255
flux for proximity fuzes, 208
techniques, 272-273

Spin effect on arming, rocket fuzes, 368

INDEX

475

Stability tests
noise sources in fuzes, 289
purpose, 287

radio-frequency sections, 283
vibration and shock production, 287-
289

Static torque tests, 299
Styramic 18, plastic for fuzes, 206
Sympathetic functioning
“of a fuze, definition, 211
of rocket fuzes, 339

T-2 arming delay device, 170, 322
T-5 fuze

see also 4.5-in. Army rocket fuzes
acceptance tests, 370-371, 431-432
amplifier, 217
applications, 213
arming, 173, 368-370
burst heights, 238, 373-375
casualties as function of burst height,
373-375

dimensions, 158

effect of distance to target, 371-372
effect of trajectory dispersion on
burst distribution, 372-373
limitation, 214
military requirements, 2
operational use, 420-422, 427
oscillator, 217

plane-to-ground firing, 216, 372-375
plane-to-plane firing, 215, 372
plastic content, 204
premature functioning, 369-370
risk of random bursts, 215
self destruction, 131, 174, 215, 217
tests, 327

zero shielding, 372-375
T-6 fuze

see also 4.5-in. Army rocket fuzes
amplifier, 217
application, 213
arming, 375-376
burst heights, 238

comparison with PD M-4 fuzes, 376
general description, 10
ground-to-ground firing, 216
impact detonator, 175
operational use, 420-422, 427
oscillator, 217
performance, summary, 13
probability of arming within certain
distance, 215
reliability, 376
T-12 fuze, 10, 177-178
T-30 fuze, 240-241, 377-384
afterburning, 380-382
arming, 186, 240-241, 377-379
characteristics, 241
compensated resistor tests, 281
effect of propellant characteristics, 380

function on approach to water, 382-
383

gear train, 186
ground-launched tests, 381
metal vane, 143
mock-plane tests, 383
plane firing, 381
plane-to-drone firing, 383-384
plane-to-water firing, 382-383
potting the amplifier, 262
power supply, 156-157
static tests in an airstream, 380-381
T-40 fuze, 3, 198
T-43 fuze, 3, 198
T-50 fuze, 179-184
adapter case, 179
air travel, 183
antenna, 89, 180
arming, 181-184
coupling, 180
design, 179-181
detonation, 183
dynamic balancing, 202
oscillator-diode circuit, 100
plastic content, 204
power supply, 156
reactor grid detector circuit, 101
self-destruction, 181
use in air-burst bombs, 412-416
vanes, 179
windmills, 179-180
T-50-E1 fuze

characteristics, 229
general description, 10
operational use, 421-423
performance, 13
plastic vane, 142
tests, 396-397
T-50-E4 fuze

characteristics, 232
general description, 10
operational use, 422-426
performance, 13
tests, 398
T-51 fuze

electronic design, 164-165
feedback amplifier, 112
general description, 10-11, 184
generator, 140
performance, 13
plastic content, 205
plastic vane, 142-143
power supply, 156
radiation resistance, 41
release, 408
RGD circuit, 101
use in air-burst bombs, 412-416
use in mustard gas bomb, 416
T-51-E1 fuze

characteristics, 234
operational use, 422-426

SECRET

performance in train, 410
release, 409
T-74 rocket, fins, 328
T-82 fuze

amplifier construction, 258
flexible blades for turbines, 201
general description, 11, 239-240
generator design, 152-153
mechanical design, 184-186
power supply, 158
release, 408, 409
turbine, 143-144
turbo-generator, 141
T-83 rocket, characteristics, 326
T-89 fuze

characteristics, 229
general description, 10
performance, 13
tests, 396-397
T-90 fuze

characteristics, 232
performance, 13
tests, 398
T-91 fuze

characteristics, 229
general description, 10
performance, 13
tests, 397

T-91-E1 fuze, tests, 397-398
T-92 fuze

characteristics, 232
general description, 10
performance, 13
T-92-E1 fuze

characteristics, 232
tests, 398

T-132 fuze, 187-195, 241-244
arming, 99, 169, 174, 190-192
arrangement of components, 189
dashpot arming device, 192
detonator rotor, 191
dynamic balancing, 190
electronic assembly, 193
end cap design, 89
features summarized, 242
general description, 11
generator design, 141, 151
generators, 189
jolt test, 191

military requirements, 3-4
oscillator, 253-256
overall dimensions, 193
performance, 417-418
plastic content, 205
power supply, 157
turbine, 144

T-171 fuze, 188-195, 241-244
arming, 99, 126, 191-192
arrangement of components, 189
detonator rotor, 191
dynamic balancing, 190

476

INDEX

end cap design, 89
generator design, 141, 151
generators, 189
jolt test, 191

military requirements, 3-4
overall dimensions, 193
performance, 418-419
plastic content, 205
potting, 207-208
power supply, 157
T-172 fuze, 195-197, 241-244
antenna, 165
general description, 11
generator design, 141, 151-152, 196
mechanical design, 195
military requirements, 3-4
nozzles for speed regulation, 200
oscillator circuit, 101-102
overall dimensions, 195
power supply, 157
T-712 bomb fuze, 234
T-2004 fuze

see also AR 5.0 Navy rocket fuze
acceptance tests, 385-386, 430-431
arming, 186, 219, 377-379
burst heights, 386
gear train, 186
general description, 11
high-angle firing, 384
metal vane, 143
performance, 13
plane-to-surface firing, 385
power supply, 156-157
T-2005 fuze
arming system, 197
general description, 11
generator design, 141
generator power supply, 197
plastic content, 206
self-destruction, 198
specification requirements, 241
Tail fuzes, 3, 198

Target, effect on impedance signal, 53-
54

Targets for proximity fuzes, 1
Telescope, battery command, 345
Temperature tests, 297
Tetryl in VT mortar fuzes, detonation,
353-356

Thermistor signal simulator, 68
Thermoplastic materials for proximity
fuzes, 204

Thermoplastic oscillator blocks, 248

Thermosetting phenolic oscillator
blocks, 248

Thyratron in detonator circuit
grid voltage, 122-123
leakage and grid current, 123
life, 123-124

low power consumption, 122
microphonics, 123
stability, 123
surge characteristics, 123
Thyratron tests, 286-287, 293
Timing devices used in arming prox-
imity fuzes, 170

Train bombing tests, VT-fuzed bombs,
322-323

Transverse-antenna fuzes, 112-113
Trench mortar fuzes

gain-frequency characteristic curve,
110

generator-powered, 165-166
T-132; 11, 187-195, 241-244
T-171; 11, 188-195, 241-244
T-172; 11, 195-197, 241-244
Tube characteristics, 90-95
diodes, 95

microphonics, 94-95
NR-3A Raytheon, 92-93
NS-3 Sylvania, 92
pentodes, 114-115

requirements and restrictions, 90-92

ruggedness, 95

self-noise, 94

testing, 290-292

triodes, 92-95

University of California

bearings for T-132 and T-171 fuzes,
190

centrifugal speed regulation for fuzes,
201

nozzles for fuzes, 200
University of Florida, T-172 fuze, 195
University of Iowa, mortar shell fuze
testing, 340-359
Uskon cloth, 281

Vacuum potting of amplifiers, 264
Vane shaft bearings for proximity fuze
noses, 265

Vibration tests, 297

Vibrators for fuze testing, 287-290

VT fuzes, 360-432

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Radio proximity fuzes for fin-stabilized missiles"

See other formats